title
stringlengths 6
244
| abstract
stringlengths 19
5.09k
| label
stringclasses 10
values |
|---|---|---|
A Drug Recommendation System (Dr.S) for cancer cell lines | Personalizing drug prescriptions in cancer care based on genomic information
requires associating genomic markers with treatment effects. This is an
unsolved challenge requiring genomic patient data in yet unavailable volumes as
well as appropriate quantitative methods. We attempt to solve this challenge
for an experimental proxy for which sufficient data is available: 42 drugs
tested on 1018 cancer cell lines. Our goal is to develop a method to identify
the drug that is most promising based on a cell line's genomic information. For
this, we need to identify for each drug the machine learning method, choice of
hyperparameters and genomic features for optimal predictive performance. We
extensively compare combinations of gene sets (both curated and random),
genetic features, and machine learning algorithms for all 42 drugs. For each
drug, the best performing combination (considering only the curated gene sets)
is selected. We use these top model parameters for each drug to build and
demonstrate a Drug Recommendation System (Dr.S). Insights resulting from this
analysis are formulated as best practices for developing drug recommendation
systems. The complete software system, called the Cell Line Analyzer, is
written in Python and available on github.
| stat |
REAK: Reliability analysis through Error rate-based Adaptive Kriging | As models in various fields are becoming more complex, associated
computational demands have been increasing significantly. Reliability analysis
for these systems when failure probabilities are small is significantly
challenging, requiring a large number of costly simulations. To address this
challenge, this paper introduces Reliability analysis through Error rate-based
Adaptive Kriging (REAK). An extension of the Central Limit Theorem based on
Lindeberg condition is adopted here to derive the distribution of the number of
design samples with wrong sign estimate and subsequently determine the maximum
error rate for failure probability estimates. This error rate enables optimal
establishment of effective sampling regions at each stage of an adaptive scheme
for strategic generation of design samples. Moreover, it facilitates setting a
target accuracy for failure probability estimation, which is used as stopping
criterion for reliability analysis. These capabilities together can
significantly reduce the number of calls to sophisticated, computationally
demanding models. The application of REAK for four examples with varying extent
of nonlinearity and dimension is presented. Results indicate that REAK is able
to reduce the computational demand by as high as 50% compared to
state-of-the-art methods of Adaptive Kriging with Monte Carlo Simulation
(AK-MCS) and Improved Sequential Kriging Reliability Analysis (ISKRA).
| stat |
Breiman's "Two Cultures" Revisited and Reconciled | In a landmark paper published in 2001, Leo Breiman described the tense
standoff between two cultures of data modeling: parametric statistical and
algorithmic machine learning. The cultural division between these two
statistical learning frameworks has been growing at a steady pace in recent
years. What is the way forward? It has become blatantly obvious that this
widening gap between "the two cultures" cannot be averted unless we find a way
to blend them into a coherent whole. This article presents a solution by
establishing a link between the two cultures. Through examples, we describe the
challenges and potential gains of this new integrated statistical thinking.
| stat |
LMLFM: Longitudinal Multi-Level Factorization Machine | We consider the problem of learning predictive models from longitudinal data,
consisting of irregularly repeated, sparse observations from a set of
individuals over time. Such data often exhibit {\em longitudinal correlation}
(LC) (correlations among observations for each individual over time), {\em
cluster correlation} (CC) (correlations among individuals that have similar
characteristics), or both. These correlations are often accounted for using
{\em mixed effects models} that include {\em fixed effects} and {\em random
effects}, where the fixed effects capture the regression parameters that are
shared by all individuals, whereas random effects capture those parameters that
vary across individuals. However, the current state-of-the-art methods are
unable to select the most predictive fixed effects and random effects from a
large number of variables, while accounting for complex correlation structure
in the data and non-linear interactions among the variables. We propose
Longitudinal Multi-Level Factorization Machine (LMLFM), to the best of our
knowledge, the first model to address these challenges in learning predictive
models from longitudinal data. We establish the convergence properties, and
analyze the computational complexity, of LMLFM. We present results of
experiments with both simulated and real-world longitudinal data which show
that LMLFM outperforms the state-of-the-art methods in terms of predictive
accuracy, variable selection ability, and scalability to data with large number
of variables. The code and supplemental material is available at
\url{https://github.com/junjieliang672/LMLFM}.
| stat |
Representing and Denoising Wearable ECG Recordings | Modern wearable devices are embedded with a range of noninvasive biomarker
sensors that hold promise for improving detection and treatment of disease. One
such sensor is the single-lead electrocardiogram (ECG) which measures
electrical signals in the heart. The benefits of the sheer volume of ECG
measurements with rich longitudinal structure made possible by wearables come
at the price of potentially noisier measurements compared to clinical ECGs,
e.g., due to movement. In this work, we develop a statistical model to simulate
a structured noise process in ECGs derived from a wearable sensor, design a
beat-to-beat representation that is conducive for analyzing variation, and
devise a factor analysis-based method to denoise the ECG. We study synthetic
data generated using a realistic ECG simulator and a structured noise model. At
varying levels of signal-to-noise, we quantitatively measure an upper bound on
performance and compare estimates from linear and non-linear models. Finally,
we apply our method to a set of ECGs collected by wearables in a mobile health
study.
| stat |
A new regret analysis for Adam-type algorithms | In this paper, we focus on a theory-practice gap for Adam and its variants
(AMSgrad, AdamNC, etc.). In practice, these algorithms are used with a constant
first-order moment parameter $\beta_{1}$ (typically between $0.9$ and $0.99$).
In theory, regret guarantees for online convex optimization require a rapidly
decaying $\beta_{1}\to0$ schedule. We show that this is an artifact of the
standard analysis and propose a novel framework that allows us to derive
optimal, data-dependent regret bounds with a constant $\beta_{1}$, without
further assumptions. We also demonstrate the flexibility of our analysis on a
wide range of different algorithms and settings.
| stat |
Continuous Hierarchical Representations with Poincar\'e Variational
Auto-Encoders | The variational auto-encoder (VAE) is a popular method for learning a
generative model and embeddings of the data. Many real datasets are
hierarchically structured. However, traditional VAEs map data in a Euclidean
latent space which cannot efficiently embed tree-like structures. Hyperbolic
spaces with negative curvature can. We therefore endow VAEs with a Poincar\'e
ball model of hyperbolic geometry as a latent space and rigorously derive the
necessary methods to work with two main Gaussian generalisations on that space.
We empirically show better generalisation to unseen data than the Euclidean
counterpart, and can qualitatively and quantitatively better recover
hierarchical structures.
| stat |
On new data sources for the production of official statistics | In the past years we have witnessed the rise of new data sources for the
potential production of official statistics, which, by and large, can be
classified as survey, administrative, and digital data. Apart from the
differences in their generation and collection, we claim that their lack of
statistical metadata, their economic value, and their lack of ownership by data
holders pose several entangled challenges lurking the incorporation of new data
into the routinely production of official statistics. We argue that every
challenge must be duly overcome in the international community to bring new
statistical products based on these sources. These challenges can be naturally
classified into different entangled issues regarding access to data,
statistical methodology, quality, information technologies, and management. We
identify the most relevant to be necessarily tackled before new data sources
can be definitively considered fully incorporated into the production of
official statistics.
| stat |
Improved GM(1,1) model based on Simpson formula and its applications | The classical GM(1,1) model is an efficient tool to {make accurate forecasts}
with limited samples. But the accuracy of the GM(1,1) model still needs to be
improved. This paper proposes a novel discrete GM(1,1) model, named ${\rm
GM_{SD}}$(1,1) model, of which the background value is reconstructed using
Simpson formula. The expression of the specific time response function is
deduced, and the relationship between our model} and the continuous GM(1,1)
model with Simpson formula called ${\rm GM_{SC} }$(1,1) model is systematically
discussed. The proposed model is proved to be unbiased to simulate the
homogeneous exponent sequence. Further, some numerical examples are given to
validate the accuracy of the new ${\rm GM_{SD}}$(1,1) model. Finally, this
model is used to predict the Gross Domestic Product and the freightage of
Lanzhou, and the results illustrate the ${\rm GM_{SD}}$(1,1) model provides
accurate prediction.
| stat |
Intuitive joint priors for variance parameters | Variance parameters in additive models are typically assigned independent
priors that do not account for model structure. We present a new framework for
prior selection based on a hierarchical decomposition of the total variance
along a tree structure to the individual model components. For each split in
the tree, an analyst may be ignorant or have a sound intuition on how to
attribute variance to the branches. In the former case a Dirichlet prior is
appropriate to use, while in the latter case a penalised complexity (PC) prior
provides robust shrinkage. A bottom-up combination of the conditional priors
results in a proper joint prior. We suggest default values for the
hyperparameters and offer intuitive statements for eliciting the
hyperparameters based on expert knowledge. The prior framework is applicable
for R packages for Bayesian inference such as INLA and RStan.
Three simulation studies show that, in terms of the application-specific
measures of interest, PC priors improve inference over Dirichlet priors when
used to penalise different levels of complexity in splits. However, when
expressing ignorance in a split, Dirichlet priors perform equally well and are
preferred for their simplicity. We find that assigning current state-of-the-art
default priors for each variance parameter individually is less transparent and
does not perform better than using the proposed joint priors. We demonstrate
practical use of the new framework by analysing spatial heterogeneity in
neonatal mortality in Kenya in 2010-2014 based on complex survey data.
| stat |
Analysis via Orthonormal Systems in Reproducing Kernel Hilbert
$C^*$-Modules and Applications | Kernel methods have been among the most popular techniques in machine
learning, where learning tasks are solved using the property of reproducing
kernel Hilbert space (RKHS). In this paper, we propose a novel data analysis
framework with reproducing kernel Hilbert $C^*$-module (RKHM), which is another
generalization of RKHS than vector-valued RKHS (vv-RKHS). Analysis with RKHMs
enables us to deal with structures among variables more explicitly than
vv-RKHS. We show the theoretical validity for the construction of orthonormal
systems in Hilbert $C^*$-modules, and derive concrete procedures for
orthonormalization in RKHMs with those theoretical properties in numerical
computations. Moreover, we apply those to generalize with RKHM kernel principal
component analysis and the analysis of dynamical systems with Perron-Frobenius
operators. The empirical performance of our methods is also investigated by
using synthetic and real-world data.
| stat |
Alternative modelling and inference methods for claim size distributions | The upper tail of a claim size distribution of a property line of business is
frequently modelled by Pareto distribution. However, the upper tail does not
need to be Pareto distributed, extraordinary shapes are possible. Here, the
opportunities for the modelling of loss distributions are extended. The basic
idea is the adjustment of a base distribution for their tails. The
(generalised) Pareto distribution is used as base distribution for different
reasons. The upper tail is in the focus and can be modelled well for special
cases by a discrete mixture of the base distribution with a combination of the
base distribution with an adapting distribution via the product of their
survival functions. A kind of smoothed step is realised in this way in the
original line function between logarithmic loss and logarithmic exceedance
probability. The lower tail can also be adjusted. The new approaches offer the
opportunity for stochastic interpretation and are applied to observed losses.
For parameter estimation, a modification of the minimum Anderson Darling
distance method is used. A new test is suggested to exclude that the observed
upper tail is better modelled by a simple Pareto distribution. Q-Q plots are
applied, and secondary results are also discussed.
| stat |
Accuracy Gains from Privacy Amplification Through Sampling for
Differential Privacy | Recent research in differential privacy demonstrated that (sub)sampling can
amplify the level of protection. For example, for $\epsilon$-differential
privacy and simple random sampling with sampling rate $r$, the actual privacy
guarantee is approximately $r\epsilon$, if a value of $\epsilon$ is used to
protect the output from the sample. In this paper, we study whether this
amplification effect can be exploited systematically to improve the accuracy of
the privatized estimate. Specifically, assuming the agency has information for
the full population, we ask under which circumstances accuracy gains could be
expected, if the privatized estimate would be computed on a random sample
instead of the full population. We find that accuracy gains can be achieved for
certain regimes. However, gains can typically only be expected, if the
sensitivity of the output with respect to small changes in the database does
not depend too strongly on the size of the database. We only focus on
algorithms that achieve differential privacy by adding noise to the final
output and illustrate the accuracy implications for two commonly used
statistics: the mean and the median. We see our research as a first step
towards understanding the conditions required for accuracy gains in practice
and we hope that these findings will stimulate further research broadening the
scope of differential privacy algorithms and outputs considered.
| stat |
Stochastic spectral embedding | Constructing approximations that can accurately mimic the behavior of complex
models at reduced computational costs is an important aspect of uncertainty
quantification. Despite their flexibility and efficiency, classical surrogate
models such as Kriging or polynomial chaos expansions tend to struggle with
highly non-linear, localized or non-stationary computational models. We hereby
propose a novel sequential adaptive surrogate modeling method based on
recursively embedding locally spectral expansions. It is achieved by means of
disjoint recursive partitioning of the input domain, which consists in
sequentially splitting the latter into smaller subdomains, and constructing a
simpler local spectral expansions in each, exploiting the trade-off complexity
vs. locality. The resulting expansion, which we refer to as "stochastic
spectral embedding" (SSE), is a piece-wise continuous approximation of the
model response that shows promising approximation capabilities, and good
scaling with both the problem dimension and the size of the training set. We
finally show how the method compares favorably against state-of-the-art sparse
polynomial chaos expansions on a set of models with different complexity and
input dimension.
| stat |
Automatic Backward Filtering Forward Guiding for Markov processes and
graphical models | We incorporate discrete and continuous time Markov processes as building
blocks into probabilistic graphical models with latent and observed variables.
We introduce the automatic Backward Filtering Forward Guiding (BFFG) paradigm
(Mider et al., 2020) for programmable inference on latent states and model
parameters. Our starting point is a generative model, a forward description of
the probabilistic process dynamics. We backpropagate the information provided
by observations through the model to transform the generative (forward) model
into a pre-conditional model guided by the data. It approximates the actual
conditional model with known likelihood-ratio between the two. The backward
filter and the forward change of measure are suitable to be incorporated into a
probabilistic programming context because they can be formulated as a set of
transformation rules.
The guided generative model can be incorporated in different approaches to
efficiently sample latent states and parameters conditional on observations. We
show applicability in a variety of settings, including Markov chains with
discrete state space, interacting particle systems, state space models,
branching diffusions and Gamma processes.
| stat |
Scientific discovery in a model-centric framework: Reproducibility,
innovation, and epistemic diversity | Consistent confirmations obtained independently of each other lend
credibility to a scientific result. We refer to results satisfying this
consistency as reproducible and assume that reproducibility is a desirable
property of scientific discovery. Yet seemingly science also progresses despite
irreproducible results, indicating that the relationship between
reproducibility and other desirable properties of scientific discovery is not
well understood. These properties include early discovery of truth, persistence
on truth once it is discovered, and time spent on truth in a long-term
scientific inquiry. We build a mathematical model of scientific discovery that
presents a viable framework to study its desirable properties including
reproducibility. In this framework, we assume that scientists adopt a
model-centric approach to discover the true model generating data in a
stochastic process of scientific discovery. We analyze the properties of this
process using Markov chain theory, Monte Carlo methods, and agent-based
modeling. We show that the scientific process may not converge to truth even if
scientific results are reproducible and that irreproducible results do not
necessarily imply untrue results. The proportion of different research
strategies represented in the scientific population, scientists' choice of
methodology, the complexity of truth, and the strength of signal contribute to
this counter-intuitive finding. Important insights include that innovative
research speeds up the discovery of scientific truth by facilitating the
exploration of model space and epistemic diversity optimizes across desirable
properties of scientific discovery.
| stat |
Irrational Exuberance: Correcting Bias in Probability Estimates | We consider the common setting where one observes probability estimates for a
large number of events, such as default risks for numerous bonds.
Unfortunately, even with unbiased estimates, selecting events corresponding to
the most extreme probabilities can result in systematically underestimating the
true level of uncertainty. We develop an empirical Bayes approach "Excess
Certainty Adjusted Probabilities" (ECAP), using a variant of Tweedie's formula,
which updates probability estimates to correct for selection bias. ECAP is a
flexible non-parametric method, which directly estimates the score function
associated with the probability estimates, so it does not need to make any
restrictive assumptions about the prior on the true probabilities. ECAP also
works well in settings where the probability estimates are biased. We
demonstrate through theoretical results, simulations, and an analysis of two
real world data sets, that ECAP can provide significant improvements over the
original probability estimates.
| stat |
Modeling Spectral Properties in Stationary Processes of Varying
Dimensions with Applications to Brain Local Field Potential Signals | A common class of methods for analyzing of multivariate time series,
stationary and nonstationary, decomposes the observed series into latent
sources. Methods such as principal compoment analysis (PCA), independent
component analysis (ICA) and Stationary Subspace Analysis (SSA) assume the
observed multivariate process is generated by latent sources that are
stationary or nonstationary. We develop a method that tracks changes in the
complexity of a 32-channel local field potential (LFP) signal from a rat
following an experimentally induced stroke. We study complexity through the
latent sources and their dimensions that can change across epochs due to an
induced shock to the cortical system. Our method compares the spread of
spectral information in several multivariate stationary processes with
different dimensions. A frequency specific spectral ratio (FS-ratio) statistic
is proposed and its asymptotic properties are derived. The FS-ratio is blind to
the dimension of the stationary process and captures the proportion of spectral
information in various (user-specified) frequency bands. We apply our method to
study differences in complexity and structure of the LFP before and after
system shock. The analysis indicates that spectral information in the beta
frequency band (12-30 Hertz) demonstrated the greatest change in structure and
complexity due to the stroke.
| stat |
Model ensembles of artificial neural networks and support vector
regression for improved accuracy in the prediction of vegetation conditions | There is increasing need for highly predictive and stable models for the
prediction of drought as an aid to better planning for drought response. This
paper presents the performance of both homogenous and heterogenous model
ensembles in the prediction of drought severity using the study case techniques
of artificial neural networks (ANN) and support vector regression (SVR). For
each of the homogenous and heterogenous model ensembles, the study investigates
the performance of three model ensembling approaches: linear averaging
(non-weighted), ranked weighted averaging and model stacking using artificial
neural networks. Using the approach of 'over-produce then select', the study
used 17 years of data on 16 selected variables for predictive drought
monitoring to build 244 individual ANN and SVR models from which 111 models
were selected for the building of the model ensembles. The results indicate
marginal superiority of heterogenous to homogenous model ensembles. Model
stacking is shown to realize models that are superior in performance in the
prediction of future vegetation conditions as compared to the linear averaging
and weighted averaging approaches. The best performance from the heterogenous
stacked model ensembles recorded an R2 of 0.94 in the prediction of future
vegetation conditions as compared to an R2 of 0.83 and R2 of 0.78 for both ANN
and SVR respectively in the traditional champion model approaches to the
realization of predictive models. We conclude that despite the computational
resource intensiveness of the model ensembling approach to drought prediction,
the returns in terms of model performance is worth the investment, especially
in the context of the recent exponential increase in computational power.
| stat |
A simulation study of methods for handling disease progression in
dose-finding clinical trials | In traditional dose-finding studies, dose-limiting toxicity (DLT) is
determined within a fixed time observation window where DLT is often defined as
a binary outcome. In the setting of oncology dose-finding trials, often
patients in advanced stage of diseases are enrolled. Therefore, disease
progression may occur within the DLT observation window leading to treatment
discontinuation and rendering the patient unevaluable for DLT assessment. As a
result, additional patients have to be enrolled, increasing the sample size. We
propose and compare several practical methods for handling disease progression
which occurs within the DLT observation window, in the context of the
time-to-event continual reassessment method (TITE-CRM) which allows using
partial observations. The methods differ on the way they define an evaluable
patient and in the way incomplete observations are included. The methods are
illustrated and contrasted in the context of a single simulated trial, and
compared via simulations under various scenarios of dose-progression
relationship, in the setting of advanced soft-tissue sarcoma.
| stat |
Teaching a GAN What Not to Learn | Generative adversarial networks (GANs) were originally envisioned as
unsupervised generative models that learn to follow a target distribution.
Variants such as conditional GANs, auxiliary-classifier GANs (ACGANs) project
GANs on to supervised and semi-supervised learning frameworks by providing
labelled data and using multi-class discriminators. In this paper, we approach
the supervised GAN problem from a different perspective, one that is motivated
by the philosophy of the famous Persian poet Rumi who said, "The art of knowing
is knowing what to ignore." In the GAN framework, we not only provide the GAN
positive data that it must learn to model, but also present it with so-called
negative samples that it must learn to avoid - we call this "The Rumi
Framework." This formulation allows the discriminator to represent the
underlying target distribution better by learning to penalize generated samples
that are undesirable - we show that this capability accelerates the learning
process of the generator. We present a reformulation of the standard GAN (SGAN)
and least-squares GAN (LSGAN) within the Rumi setting. The advantage of the
reformulation is demonstrated by means of experiments conducted on MNIST,
Fashion MNIST, CelebA, and CIFAR-10 datasets. Finally, we consider an
application of the proposed formulation to address the important problem of
learning an under-represented class in an unbalanced dataset. The Rumi approach
results in substantially lower FID scores than the standard GAN frameworks
while possessing better generalization capability.
| stat |
Prediction in the presence of response-dependent missing labels | In a variety of settings, limitations of sensing technologies or other
sampling mechanisms result in missing labels, where the likelihood of a missing
label in the training set is an unknown function of the data. For example,
satellites used to detect forest fires cannot sense fires below a certain size
threshold. In such cases, training datasets consist of positive and
pseudo-negative observations where pseudo-negative observations can be either
true negatives or undetected positives with small magnitudes. We develop a new
methodology and non-convex algorithm P(ositive) U(nlabeled) - O(ccurrence)
M(agnitude) M(ixture) which jointly estimates the occurrence and detection
likelihood of positive samples, utilizing prior knowledge of the detection
mechanism. Our approach uses ideas from positive-unlabeled (PU)-learning and
zero-inflated models that jointly estimate the magnitude and occurrence of
events. We provide conditions under which our model is identifiable and prove
that even though our approach leads to a non-convex objective, any local
minimizer has optimal statistical error (up to a log term) and projected
gradient descent has geometric convergence rates. We demonstrate on both
synthetic data and a California wildfire dataset that our method out-performs
existing state-of-the-art approaches.
| stat |
Matrix Completion using Kronecker Product Approximation | A matrix completion problem is to recover the missing entries in a partially
observed matrix. Most of the existing matrix completion methods assume a low
rank structure of the underlying complete matrix. In this paper, we introduce
an alternative and more general form of the underlying complete matrix, which
assumes a low Kronecker rank instead of a low regular rank, but includes the
latter as a special case. The extra flexibility allows for a much more
parsimonious representation of the underlying matrix, but also raises the
challenge of determining the proper Kronecker product configuration to be used.
We find that the configuration can be identified using the mean squared error
criterion as well as a modified cross-validation criterion. We establish the
consistency of this procedure under suitable conditions on the signal-to-noise
ratio. A aggregation procedure is also proposed to deal with special missing
patterns and complex underlying structures. Both numerical and empirical
studies are carried out to demonstrate the performance of the new method.
| stat |
Testing independence between two random sets for the analysis of
colocalization in bio-imaging | Colocalization aims at characterizing spatial associations between two
fluorescently-tagged biomolecules by quantifying the co-occurrence and
correlation between the two channels acquired in fluorescence microscopy.
Colocalization is presented either as the degree of overlap between the two
channels or the overlays of the red and green images, with areas of yellow
indicating colocalization of the molecules. This problem remains an open issue
in diffraction-limited microscopy and raises new challenges with the emergence
of super-resolution imaging, a microscopic technique awarded by the 2014 Nobel
prize in chemistry. We propose GcoPS, for Geo-coPositioning System, an original
method that exploits the random sets structure of the tagged molecules to
provide an explicit testing procedure. Our simulation study shows that GcoPS
unequivocally outperforms the best competitive methods in adverse situations
(noise, irregularly shaped fluorescent patterns, different optical
resolutions). GcoPS is also much faster, a decisive advantage to face the huge
amount of data in super-resolution imaging. We demonstrate the performances of
GcoPS on two biological real datasets, obtained by conventional
diffraction-limited microscopy technique and by super-resolution technique,
respectively.
| stat |
Stochastic Properties of Minimal Arc Distance and Cosine Similarity
between a Random Point and Prespecified Sites on Sphere | In applications such as wireless communication, it is important to study the
statistical properties of $L_{2}$, the minimal arc distance between a random
point (e.g., a cellphone user) uniformly distributed on a sphere to a set of
pre-defined seeds (e.g., wireless towers) on that sphere. In this study, we
first derive the distribution (CDF) and density (PDF) functions of the arc
distance between a selected vertex of a spherical triangle to a random point
uniformly distributed within this triangle. Next, using computational
techniques based on spherical Voronoi diagram and triangular partition of
Voronoi cells, we derive moments of $L_{2}$ and $\cos L_{2}$. These results are
verified by extensive Monte Carlo simulations.
| stat |
Bayesian cluster analysis: Point estimation and credible balls | Clustering is widely studied in statistics and machine learning, with
applications in a variety of fields. As opposed to classical algorithms which
return a single clustering solution, Bayesian nonparametric models provide a
posterior over the entire space of partitions, allowing one to assess
statistical properties, such as uncertainty on the number of clusters. However,
an important problem is how to summarize the posterior; the huge dimension of
partition space and difficulties in visualizing it add to this problem. In a
Bayesian analysis, the posterior of a real-valued parameter of interest is
often summarized by reporting a point estimate such as the posterior mean along
with 95% credible intervals to characterize uncertainty. In this paper, we
extend these ideas to develop appropriate point estimates and credible sets to
summarize the posterior of clustering structure based on decision and
information theoretic techniques.
| stat |
Modelling columnarity of pyramidal cells in the human cerebral cortex | For modelling the location of pyramidal cells in the human cerebral cortex we
suggest a hierarchical point process in $\mathbb{R}^3$ that exhibits anisotropy
in the form of cylinders extending along the $z$-axis. The model consists first
of a generalised shot noise Cox process for the $xy$-coordinates, providing
cylindrical clusters, and next of a Markov random field model for the
$z$-coordinates conditioned on the $xy$-coordinates, providing either
repulsion, aggregation, or both within specified areas of interaction. Several
cases of these hierarchical point processes are fitted to two pyramidal cell
datasets, and of these a final model allowing for both repulsion and attraction
between the points seem adequate. We discuss how the final model relates to the
so-called minicolumn hypothesis in neuroscience.
| stat |
Testing marginal homogeneity in Hilbert spaces with applications to
stock market returns | The paper considers a paired data framework and discuss the question of
marginal homogeneity of bivariate high dimensional or functional data. The
related testing problem can be endowed into a more general setting for paired
random variables taking values in a general Hilbert space. To address this
problem, a Cramer-von-Mises type test statistic is applied and a bootstrap
procedure is suggested to obtain critical values and finally a consistent test.
The desired properties of a bootstrap test can be derived, that are asymptotic
exactness under the null hypothesis and consistency under alternatives.
Simulations show the quality of the test in the finite sample case. A possible
application is the comparison of two possibly dependent stock market returns on
the basis of functional data. The approach is demonstrated on the basis of
historical data for different stock market indices.
| stat |
Robust Risk Minimization for Statistical Learning | We consider a general statistical learning problem where an unknown fraction
of the training data is corrupted. We develop a robust learning method that
only requires specifying an upper bound on the corrupted data fraction. The
method minimizes a risk function defined by a non-parametric distribution with
unknown probability weights. We derive and analyse the optimal weights and show
how they provide robustness against corrupted data. Furthermore, we give a
computationally efficient coordinate descent algorithm to solve the risk
minimization problem. We demonstrate the wide range applicability of the
method, including regression, classification, unsupervised learning and classic
parameter estimation, with state-of-the-art performance.
| stat |
Modelling Heterogeneity Using Bayesian Structured Sparsity | How to estimate heterogeneity, e.g. the effect of some variable differing
across observations, is a key question in political science. Methods for doing
so make simplifying assumptions about the underlying nature of the
heterogeneity to draw reliable inferences. This paper allows a common way of
simplifying complex phenomenon (placing observations with similar effects into
discrete groups) to be integrated into regression analysis. The framework
allows researchers to (i) use their prior knowledge to guide which groups are
permissible and (ii) appropriately quantify uncertainty. The paper does this by
extending work on "structured sparsity" from a traditional penalized likelihood
approach to a Bayesian one by deriving new theoretical results and inferential
techniques. It shows that this method outperforms state-of-the-art methods for
estimating heterogeneous effects when the underlying heterogeneity is grouped
and more effectively identifies groups of observations with different effects
in observational data.
| stat |
Efficient Principally Stratified Treatment Effect Estimation in
Crossover Studies with Absorbent Binary Endpoints | Suppose one wishes to estimate the effect of a binary treatment on a binary
endpoint conditional on a post-randomization quantity in a counterfactual world
in which all subjects received treatment. It is generally difficult to identify
this parameter without strong, untestable assumptions. It has been shown that
identifiability assumptions become much weaker under a crossover design in
which subjects not receiving treatment are later given treatment. Under the
assumption that the post-treatment biomarker observed in these crossover
subjects is the same as would have been observed had they received treatment at
the start of the study, one can identify the treatment effect with only mild
additional assumptions. This remains true if the endpoint is absorbent, i.e. an
endpoint such as death or HIV infection such that the post-crossover treatment
biomarker is not meaningful if the endpoint has already occurred. In this work,
we review identifiability results for a parameter of the distribution of the
data observed under a crossover design with the principally stratified
treatment effect of interest. We describe situations in which these assumptions
would be falsifiable, and show that these assumptions are not otherwise
falsifiable. We then provide a targeted minimum loss-based estimator for the
setting that makes no assumptions on the distribution that generated the data.
When the semiparametric efficiency bound is well defined, for which the primary
condition is that the biomarker is discrete-valued, this estimator is efficient
among all regular and asymptotically linear estimators. We also present a
version of this estimator for situations in which the biomarker is continuous.
Implications to closeout designs for vaccine trials are discussed.
| stat |
Intensity Estimation on Geometric Networks with Penalized Splines | In the past decades, the growing amount of network data has lead to many
novel statistical models. In this paper we consider so called geometric
networks. Typical examples are road networks or other infrastructure networks.
But also the neurons or the blood vessels in a human body can be interpreted as
a geometric network embedded in a three-dimensional space. In all these
applications a network specific metric rather than the Euclidean metric is
usually used, which makes the analyses on network data challenging. We consider
network based point processes and our task is to estimate the intensity (or
density) of the process which allows to detect high- and low- intensity regions
of the underlying stochastic processes. Available routines that tackle this
problem are commonly based on kernel smoothing methods. However, kernel based
estimation in general exhibits some drawbacks such as suffering from boundary
effects and the locality of the smoother. In an Euclidean space, the
disadvantages of kernel methods can be overcome by using penalized spline
smoothing. We here extend penalized spline smoothing towards smooth intensity
estimation on geometric networks and apply the approach to both, simulated and
real world data. The results show that penalized spline based intensity
estimation is numerically efficient and outperforms kernel based methods.
Furthermore, our approach easily allows to incorporate covariates, which allows
to respect the network geometry in a regression model framework.
| stat |
Divide-and-conquer methods for big data analysis | In the context of big data analysis, the divide-and-conquer methodology
refers to a multiple-step process: first splitting a data set into several
smaller ones; then analyzing each set separately; finally combining results
from each analysis together. This approach is effective in handling large data
sets that are unsuitable to be analyzed entirely by a single computer due to
limits either from memory storage or computational time. The combined results
will provide a statistical inference which is similar to the one from analyzing
the entire data set. This article reviews some recently developments of
divide-and-conquer methods in a variety of settings, including combining based
on parametric, semiparametric and nonparametric models, online sequential
updating methods, among others. Theoretical development on the efficiency of
the divide-and-conquer methods is also discussed.
| stat |
A Review of Spatiotemporal Models for Count Data in R Packages. A Case
Study of COVID-19 Data | Spatio-temporal models for count data are required in a wide range of
scientific fields and they have become particularly crucial nowadays because of
their ability to analyse COVID-19-related data. Models for count data are
needed when the variable of interest take only non-negative integer values and
these integers arise from counting occurrences. Several R-packages are
currently available to deal with spatiotemporal areal count data. Each package
focuses on different models and/or statistical methodologies. Unfortunately,
the results generated by these models are rarely comparable due to differences
in notation and methods. The main objective of this paper is to present a
review describing the most important approaches that can be used to model and
analyse count data when questions of scientific interest concern both their
spatial and their temporal behaviour and we monitor their performance under the
same data set.
For this review, we focus on the three R-packages that can be used for this
purpose and the different models assessed are representative of the two most
widespread methodologies used to analyse spatiotemporal count data: the
classical approach (based on Penalised Likelihood or Estimating Equations) and
the Bayesian point of view.
A case study is analysed as an illustration of these different methodologies.
In this case study, these packages are used to model and predict daily
hospitalisations from COVID-19 in 24 health regions within the Valencian
Community (Spain), with data corresponding to the period from 28 June to 13
December 2020. Because of the current urgent need for monitoring and predicting
data in the COVID-19 pandemic, this case study is, in itself, of particular
importance and can be considered the secondary objective of this work.
Satisfactory and promising results have been obtained in this second goal.
| stat |
Towards new cross-validation-based estimators for Gaussian process
regression: efficient adjoint computation of gradients | We consider the problem of estimating the parameters of the covariance
function of a Gaussian process by cross-validation. We suggest using new
cross-validation criteria derived from the literature of scoring rules. We also
provide an efficient method for computing the gradient of a cross-validation
criterion. To the best of our knowledge, our method is more efficient than what
has been proposed in the literature so far. It makes it possible to lower the
complexity of jointly evaluating leave-one-out criteria and their gradients.
| stat |
An analysis of European crash data and scenario specification for heavy
truck safety system development within the AEROFLEX project | Heavy goods vehicles (HGVs) are involved in 4.5% of police-reported road
crashes in Europe and 14.2% of fatal road crashes. Active and passive safety
systems can help to prevent crashes or mitigate the consequences but need
detailed scenarios to be designed effectively. The aim of this paper is to give
a comprehensive and up-to-date analysis of HGV crashes in Europe. The analysis
is based on general statistics from CARE, results about trucks weighing 16 tons
or more from national crash databases and a detailed study of in-depth crash
data from GIDAS. Three scenarios are identified that should be addressed by
future safety systems: (1) rear-end crashes with other vehicles in which the
truck is the striking partner, (2) conflicts during right turn maneuvers of the
truck and a cyclist and (3) pedestrians crossing the road perpendicular to the
direction of travel of the truck.
| stat |
Block what you can, except when you shouldn't | Several branches of the potential outcome causal inference literature have
discussed the merits of blocking versus complete randomization. Some have
concluded it can never hurt the precision of estimates, and some have concluded
it can hurt. In this paper, we reconcile these apparently conflicting views,
give a more thorough discussion of what guarantees no harm, and discuss how
other aspects of a blocked design can cost, all in terms of precision. We
discuss how the different findings are due to different sampling models and
assumptions of how the blocks were formed. We also connect these ideas to
common misconceptions, for instance showing that analyzing a blocked experiment
as if it were completely randomized, a seemingly conservative method, can
actually backfire in some cases. Overall, we find that blocking can have a
price, but that this price is usually small and the potential for gain can be
large. It is hard to go too far wrong with blocking.
| stat |
Comparison of Persistence Diagrams | Topological Data Analysis (TDA) is an approach to handle with big data by
studying its shape. A main tool of TDA is the persistence diagram, and one can
use it to compare data sets. One approach to learn on the similarity between
two persistence diagrams is to use the Bottleneck and the Wasserstein
distances. Another approach is to fit a parametric model for each diagram, and
then to compare the model coefficients. We study the behaviour of both distance
measures and the RST parametric model. The theoretical behaviour of the
distance measures is difficult to be developed, and therefore we study their
behaviour numerically. We conclude that the RST model has an advantage over the
Bottleneck and the Wasserstein distances in sense that it can give a definite
conclusion regarding the similarity between two persistence diagrams. More of
that, a great advantage of the RST is its ability to distinguish between two
data sets that are geometrically different but topologically are the same,
which is impossible to have by the two distance measures.
| stat |
Informed Bayesian Inference for the A/B Test | Booming in business and a staple analysis in medical trials, the A/B test
assesses the effect of an intervention or treatment by comparing its success
rate with that of a control condition. Across many practical applications, it
is desirable that (1) evidence can be obtained in favor of the null hypothesis
that the treatment is ineffective; (2) evidence can be monitored as the data
accumulate; (3) expert prior knowledge can be taken into account. Most existing
approaches do not fulfill these desiderata. Here we describe a Bayesian A/B
procedure based on Kass and Vaidyanathan (1992) that allows one to monitor the
evidence for the hypotheses that the treatment has either a positive effect, a
negative effect, or, crucially, no effect. Furthermore, this approach enables
one to incorporate expert knowledge about the relative prior plausibility of
the rival hypotheses and about the expected size of the effect, given that it
is non-zero. To facilitate the wider adoption of this Bayesian procedure we
developed the abtest package in R. We illustrate the package options and the
associated statistical results with a fictitious business example and a real
data medical example.
| stat |
Adaptive Smoothing Spline Estimator for the Function-on-Function Linear
Regression Model | In this paper, we propose an adaptive smoothing spline (AdaSS) estimator for
the function-on-function linear regression model where each value of the
response, at any domain point, depends on the full trajectory of the predictor.
The AdaSS estimator is obtained by the optimization of an objective function
with two spatially adaptive penalties, based on initial estimates of the
partial derivatives of the regression coefficient function. This allows the
proposed estimator to adapt more easily to the true coefficient function over
regions of large curvature and not to be undersmoothed over the remaining part
of the domain. A novel evolutionary algorithm is developed ad hoc to obtain the
optimization tuning parameters. Extensive Monte Carlo simulations have been
carried out to compare the AdaSS estimator with competitors that have already
appeared in the literature before. The results show that our proposal mostly
outperforms the competitor in terms of estimation and prediction accuracy.
Lastly, those advantages are illustrated also on two real-data benchmark
examples.
| stat |
Analysis and Design of Thompson Sampling for Stochastic Partial
Monitoring | We investigate finite stochastic partial monitoring, which is a general model
for sequential learning with limited feedback. While Thompson sampling is one
of the most promising algorithms on a variety of online decision-making
problems, its properties for stochastic partial monitoring have not been
theoretically investigated, and the existing algorithm relies on a heuristic
approximation of the posterior distribution. To mitigate these problems, we
present a novel Thompson-sampling-based algorithm, which enables us to exactly
sample the target parameter from the posterior distribution. Besides, we prove
that the new algorithm achieves the logarithmic problem-dependent expected
pseudo-regret $\mathrm{O}(\log T)$ for a linearized variant of the problem with
local observability. This result is the first regret bound of Thompson sampling
for partial monitoring, which also becomes the first logarithmic regret bound
of Thompson sampling for linear bandits.
| stat |
A Comparative Study of Parametric Regression Models to Detect Breakpoint
in Traffic Fundamental Diagram | A speed threshold is a crucial parameter in breakdown and capacity
distribution analysis as it defines the boundary between free-flow and
congested regimes. However, literature on approaches to establishing the
breakpoint value for detecting breakdown events is limited. Most of existing
studies rely on the use of either visual observation or predefined thresholds.
These approaches may not be reliable considering the variations associated with
field data. Thus, this study compared the performance of two data-driven
methods, that is, logistic function (LGF) and two-regime models, used to
establish the breakpoint from traffic flow variables. The two models were
calibrated using urban freeway traffic data. The models'performance results
revealed that with less computation efforts, the LGF has slightly better
prediction accuracy than the two-regime model. Although the two-regime model
had relatively lower performance, it can be useful in identifying the
transitional state.
| stat |
Forecast score distributions with imperfect observations | The classical paradigm of scoring rules is to discriminate between two
different forecasts by comparing them with observations. The probability
distribution of the observed record is assumed to be perfect as a verification
benchmark. In practice, however, observations are almost always tainted by
errors and uncertainties. If the yardstick used to compare forecasts is
imprecise, one can wonder whether such types of errors may or may not have a
strong influence on decisions based on classical scoring rules. We propose a
new scoring rule scheme in the context of models that incorporate errors of the
verification data. We rely on existing scoring rules and incorporate
uncertainty and error of the verification data through a hidden variable and
the conditional expectation of scores when they are viewed as a random
variable. The proposed scoring framework is compared to scores used in
practice, and is expressed in various setups, mainly an additive Gaussian noise
model and a multiplicative Gamma noise model. By considering scores as random
variables one can access the entire range of their distribution. In particular
we illustrate that the commonly used mean score can be a misleading
representative of the distribution when this latter is highly skewed or have
heavy tails. In a simulation study, through the power of a statistical test and
the computation of Wasserstein distances between scores distributions, we
demonstrate the ability of the newly proposed score to better discriminate
between forecasts when verification data are subject to uncertainty compared
with the scores used in practice. Finally, we illustrate the benefit of
accounting for the uncertainty of the verification data into the scoring
procedure on a dataset of surface wind speed from measurements and numerical
model outputs.
| stat |
Divide-and-Conquer Information-Based Optimal Subdata Selection Algorithm | The information-based optimal subdata selection (IBOSS) is a computationally
efficient method to select informative data points from large data sets through
processing full data by columns. However, when the volume of a data set is too
large to be processed in the available memory of a machine, it is infeasible to
implement the IBOSS procedure. This paper develops a divide-and-conquer IBOSS
approach to solving this problem, in which the full data set is divided into
smaller partitions to be loaded into the memory and then subsets of data are
selected from each partitions using the IBOSS algorithm. We derive both finite
sample properties and asymptotic properties of the resulting estimator.
Asymptotic results show that if the full data set is partitioned randomly and
the number of partitions is not very large, then the resultant estimator has
the same estimation efficiency as the original IBOSS estimator. We also carry
out numerical experiments to evaluate the empirical performance of the proposed
method.
| stat |
Particle Gibbs Sampling for Bayesian Phylogenetic inference | The combinatorial sequential Monte Carlo (CSMC) has been demonstrated to be
an efficient complementary method to the standard Markov chain Monte Carlo
(MCMC) for Bayesian phylogenetic tree inference using biological sequences. It
is appealing to combine the CSMC and MCMC in the framework of the particle
Gibbs (PG) sampler to jointly estimate the phylogenetic trees and evolutionary
parameters. However, the Markov chain of the particle Gibbs may mix poorly if
the underlying SMC suffers from the path degeneracy issue. Some remedies,
including the particle Gibbs with ancestor sampling and the interacting
particle MCMC, have been proposed to improve the PG. But they either cannot be
applied to or remain inefficient for the combinatorial tree space. We introduce
a novel CSMC method by proposing a more efficient proposal distribution. It
also can be combined into the particle Gibbs sampler framework to infer
parameters in the evolutionary model. The new algorithm can be easily
parallelized by allocating samples over different computing cores. We validate
that the developed CSMC can sample trees more efficiently in various particle
Gibbs samplers via numerical experiments. Our implementation is available at
https://github.com/liangliangwangsfu/phyloPMCMC
| stat |
Sequential change-point detection for mutually exciting point processes
over networks | We present a new CUSUM procedure for sequentially detecting change-point in
the self and mutual exciting processes, a.k.a. Hawkes networks using discrete
events data. Hawkes networks have become a popular model for statistics and
machine learning due to their capability in modeling irregularly observed data
where the timing between events carries a lot of information. The problem of
detecting abrupt changes in Hawkes networks arises from various applications,
including neuronal imaging, sensor network, and social network monitoring.
Despite this, there has not been a computationally and memory-efficient online
algorithm for detecting such changes from sequential data. We present an
efficient online recursive implementation of the CUSUM statistic for Hawkes
processes, both decentralized and memory-efficient, and establish the
theoretical properties of this new CUSUM procedure. We then show that the
proposed CUSUM method achieves better performance than existing methods,
including the Shewhart procedure based on count data, the generalized
likelihood ratio (GLR) in the existing literature, and the standard score
statistic. We demonstrate this via a simulated example and an application to
population code change-detection in neuronal networks.
| stat |
Nonasymptotic estimation and support recovery for high dimensional
sparse covariance matrices | We propose a general framework for nonasymptotic covariance matrix estimation
making use of concentration inequality-based confidence sets. We specify this
framework for the estimation of large sparse covariance matrices through
incorporation of past thresholding estimators with key emphasis on support
recovery. This technique goes beyond past results for thresholding estimators
by allowing for a wide range of distributional assumptions beyond merely
sub-Gaussian tails. This methodology can furthermore be adapted to a wide range
of other estimators and settings. The usage of nonasymptotic dimension-free
confidence sets yields good theoretical performance. Through extensive
simulations, it is demonstrated to have superior performance when compared with
other such methods. In the context of support recovery, we are able to specify
a false positive rate and optimize to maximize the true recoveries.
| stat |
Factorization Machines with Regularization for Sparse Feature
Interactions | Factorization machines (FMs) are machine learning predictive models based on
second-order feature interactions and FMs with sparse regularization are called
sparse FMs. Such regularizations enable feature selection, which selects the
most relevant features for accurate prediction, and therefore they can
contribute to the improvement of the model accuracy and interpretability.
However, because FMs use second-order feature interactions, the selection of
features often causes the loss of many relevant feature interactions in the
resultant models. In such cases, FMs with regularization specially designed for
feature interaction selection trying to achieve interaction-level sparsity may
be preferred instead of those just for feature selection trying to achieve
feature-level sparsity. In this paper, we present a new regularization scheme
for feature interaction selection in FMs. The proposed regularizer is an upper
bound of the $\ell_1$ regularizer for the feature interaction matrix, which is
computed from the parameter matrix of FMs. For feature interaction selection,
our proposed regularizer makes the feature interaction matrix sparse without a
restriction on sparsity patterns imposed by the existing methods. We also
describe efficient proximal algorithms for the proposed FMs and present
theoretical analyses of both existing and the new regularize. In addition, we
will discuss how our ideas can be applied or extended to more accurate feature
selection and other related models such as higher-order FMs and the all-subsets
model. The analysis and experimental results on synthetic and real-world
datasets show the effectiveness of the proposed methods.
| stat |
Semidefinite Programming for Community Detection with Side Information | This paper produces an efficient Semidefinite Programming (SDP) solution for
community detection that incorporates non-graph data, which in this context is
known as side information. SDP is an efficient solution for standard community
detection on graphs. We formulate a semi-definite relaxation for the maximum
likelihood estimation of node labels, subject to observing both graph and
non-graph data. This formulation is distinct from the SDP solution of standard
community detection, but maintains its desirable properties. We calculate the
exact recovery threshold for three types of non-graph information, which in
this paper are called side information: partially revealed labels, noisy
labels, as well as multiple observations (features) per node with arbitrary but
finite cardinality. We find that SDP has the same exact recovery threshold in
the presence of side information as maximum likelihood with side information.
Thus, the methods developed herein are computationally efficient as well as
asymptotically accurate for the solution of community detection in the presence
of side information. Simulations show that the asymptotic results of this paper
can also shed light on the performance of SDP for graphs of modest size.
| stat |
Exact Sampling of Determinantal Point Processes without
Eigendecomposition | Determinantal point processes (DPPs) enable the modeling of repulsion: they
provide diverse sets of points. The repulsion is encoded in a kernel $K$ that
can be seen as a matrix storing the similarity between points. The diversity
comes from the fact that the inclusion probability of a subset is equal to the
determinant of a submatrice of $K$. The exact algorithm to sample DPPs uses the
spectral decomposition of $K$, a computation that becomes costly when dealing
with a high number of points. Here, we present an alternative exact algorithm
in the discrete setting that avoids the eigenvalues and the eigenvectors
computation. Instead, it relies on Cholesky decompositions. This is a two steps
strategy: first, it samples a Bernoulli point process with an appropriate
distribution, then it samples the target DPP distribution through a thinning
procedure. Not only is the method used here innovative, but this algorithm can
be competitive with the original algorithm or even faster for some applications
specified here.
| stat |
Kernels over Sets of Finite Sets using RKHS Embeddings, with Application
to Bayesian (Combinatorial) Optimization | We focus on kernel methods for set-valued inputs and their application to
Bayesian set optimization, notably combinatorial optimization. We investigate
two classes of set kernels that both rely on Reproducing Kernel Hilbert Space
embeddings, namely the ``Double Sum'' (DS) kernels recently considered in
Bayesian set optimization, and a class introduced here called ``Deep
Embedding'' (DE) kernels that essentially consists in applying a radial kernel
on Hilbert space on top of the canonical distance induced by another kernel
such as a DS kernel. We establish in particular that while DS kernels typically
suffer from a lack of strict positive definiteness, vast subclasses of DE
kernels built upon DS kernels do possess this property, enabling in turn
combinatorial optimization without requiring to introduce a jitter parameter.
Proofs of theoretical results about considered kernels are complemented by a
few practicalities regarding hyperparameter fitting. We furthermore demonstrate
the applicability of our approach in prediction and optimization tasks, relying
both on toy examples and on two test cases from mechanical engineering and
hydrogeology, respectively. Experimental results highlight the applicability
and compared merits of the considered approaches while opening new perspectives
in prediction and sequential design with set inputs.
| stat |
Change point analysis in non-stationary processes - a mass excess
approach | This paper considers the problem of testing if a sequence of means
$(\mu_t)_{t =1,\ldots ,n }$ of a non-stationary time series $(X_t)_{t =1,\ldots
,n }$ is stable in the sense that the difference of the means $\mu_1$ and
$\mu_t$ between the initial time $t=1$ and any other time is smaller than a
given level, that is $ | \mu_1 - \mu_t | \leq c $ for all $t =1,\ldots ,n $. A
test for hypotheses of this type is developed using a biascorrected monotone
rearranged local linear estimator and asymptotic normality of the corresponding
test statistic is established. As the asymptotic variance depends on the
location and order of the critical roots of the equation $| \mu_1 - \mu_t | =
c$ a new bootstrap procedure is proposed to obtain critical values and its
consistency is established. As a consequence we are able to quantitatively
describe relevant deviations of a non-stationary sequence from its initial
value. The results are illustrated by means of a simulation study and by
analyzing data examples.
| stat |
Residual Bootstrap Exploration for Bandit Algorithms | In this paper, we propose a novel perturbation-based exploration method in
bandit algorithms with bounded or unbounded rewards, called residual bootstrap
exploration (\texttt{ReBoot}). The \texttt{ReBoot} enforces exploration by
injecting data-driven randomness through a residual-based perturbation
mechanism. This novel mechanism captures the underlying distributional
properties of fitting errors, and more importantly boosts exploration to escape
from suboptimal solutions (for small sample sizes) by inflating variance level
in an \textit{unconventional} way. In theory, with appropriate variance
inflation level, \texttt{ReBoot} provably secures instance-dependent
logarithmic regret in Gaussian multi-armed bandits. We evaluate the
\texttt{ReBoot} in different synthetic multi-armed bandits problems and observe
that the \texttt{ReBoot} performs better for unbounded rewards and more
robustly than \texttt{Giro} \cite{kveton2018garbage} and \texttt{PHE}
\cite{kveton2019perturbed}, with comparable computational efficiency to the
Thompson sampling method.
| stat |
Estimation of group means in generalized linear mixed models | In this manuscript, we investigate the concept of the mean response for a
treatment group mean as well as its estimation and prediction for generalized
linear models with a subject-wise random effect. Generalized linear models are
commonly used to analyze categorical data. The model-based mean for a treatment
group usually estimates the response at the mean covariate. However, the mean
response for the treatment group for studied population is at least equally
important in the context of clinical trials. New methods were proposed to
estimate such a mean response in generalized linear models; however, this has
only been done when there are no random effects in the model. We suggest that,
in a generalized linear mixed model (GLMM), there are at least two possible
definitions of a treatment group mean response that can serve as
estimation/prediction targets. The estimation of these treatment group means is
important for healthcare professionals to be able to understand the absolute
benefit versus risk. For both of these treatment group means, we propose a new
set of methods that suggests how to estimate/predict both of them in a GLMM
models with a univariate subject-wise random effect. Our methods also suggest
an easy way of constructing corresponding confidence and prediction intervals
for both possible treatment group means. Simulations show that proposed
confidence and prediction intervals provide correct empirical coverage
probability under most circumstances. Proposed methods have also been applied
to analyze hypoglycemia data from diabetes clinical trials.
| stat |
Examining socioeconomic factors to understand the hospital case-fatality
rates of COVID-19 in the city of Sao Paulo, Brazil | Understanding differences in hospital case-fatality rates (HCFRs) of
coronavirus disease (COVID-19) may help evaluate its severity and the capacity
of the healthcare system to reduce mortality. We examined the variability in
HCFRs of COVID-19 in relation to spatial inequalities in socioeconomic factors
across the city of Sao Paulo, Brazil. We found that HCFRs were higher for men
and for individuals aged 60 years and older. Our models identified per capita
income as a significant factor that is negatively associated with the HCFRs of
COVID-19, even after adjusting by age, sex and presence of risk factors.
| stat |
Regressor: A C program for Combinatorial Regressions | In statistics, researchers use Regression models for data analysis and
prediction in many productive sectors (industry, business, academy, etc.).
Regression models are mathematical functions representing an approximation of
dependent variable $Y$ from n independent variables $X_i \in X$. The literature
presents many regression methods divided into single and multiple regressions.
There are several procedures to generate regression models and sets of
commercial and academic tools that implement these procedures. This work
presents one open-source program called Regressor that makes models from a
specific variation of polynomial regression. These models relate the
independent variables to generate an approximation of the original output
dependent data. In many tests, Regressor was able to build models five times
more accurate than commercial tools.
| stat |
Multitasking additional-to-driving: Prevalence, structure, and
associated risk in SHRP2 naturalistic driving data | This paper 1) analyzes the extent to which drivers engage in multitasking
additional-to-driving (MAD) under various conditions, 2) specifies odds ratios
(ORs) of crashing associated with MAD compared to no task engagement, and 3)
explores the structure of MAD, based on data from the Second Strategic Highway
Research Program Naturalistic Driving Study (SHRP2 NDS). Sensitivity analysis
in which secondary tasks were re-defined by grouping similar tasks was
performed to investigate the extent to which ORs are affected by the specific
task definitions in SHRP2. A novel visual representation of multitasking was
developed to show which secondary tasks co-occur frequently and which ones do
not. MAD occurs in 11% of control driving segments, 22% of crashes and
near-crashes (CNC), 26% of Level 1-3 crashes and 39% of rear-end striking
crashes, and 9%, 16%, 17% and 28% respectively for the same event types if MAD
is defined in terms of general task groups. The most common co-occurrences of
secondary tasks vary substantially among event types; for example, 'Passenger
in adjacent seat - interaction' and 'Other non-specific internal eye glance'
tend to co-occur in CNC but tend not to co-occur in control driving segments.
The odds ratios of MAD compared to driving without any secondary task and the
corresponding 95% confidence intervals are 2.38 (2.17-2.61) for CNC, 3.72
(3.11-4.45) for Level 1-3 crashes and 8.48 (5.11-14.07) for rear-end striking
crashes. The corresponding ORs using general task groups to define MAD are
slightly lower at 2.00 (1.80-2.21) for CNC, 3.03 (2.48-3.69) for Level 1-3
crashes and 6.94 (4.04-11.94) for rear-end striking crashes. The results
confirm that independently of whether secondary tasks are defined according to
SHRP2 or general task groups, the reduction of driving performance from MAD
observed in simulator studies is manifested in real-world crashes as well.
| stat |
Selecting Data Augmentation for Simulating Interventions | Machine learning models trained with purely observational data and the
principle of empirical risk minimization \citep{vapnik_principles_1992} can
fail to generalize to unseen domains. In this paper, we focus on the case where
the problem arises through spurious correlation between the observed domains
and the actual task labels. We find that many domain generalization methods do
not explicitly take this spurious correlation into account. Instead, especially
in more application-oriented research areas like medical imaging or robotics,
data augmentation techniques that are based on heuristics are used to learn
domain invariant features. To bridge the gap between theory and practice, we
develop a causal perspective on the problem of domain generalization. We argue
that causal concepts can be used to explain the success of data augmentation by
describing how they can weaken the spurious correlation between the observed
domains and the task labels. We demonstrate that data augmentation can serve as
a tool for simulating interventional data. We use these theoretical insights to
derive a simple algorithm that is able to select data augmentation techniques
that will lead to better domain generalization.
| stat |
XBART: Accelerated Bayesian Additive Regression Trees | Bayesian additive regression trees (BART) (Chipman et. al., 2010) is a
powerful predictive model that often outperforms alternative models at
out-of-sample prediction. BART is especially well-suited to settings with
unstructured predictor variables and substantial sources of unmeasured
variation as is typical in the social, behavioral and health sciences. This
paper develops a modified version of BART that is amenable to fast posterior
estimation. We present a stochastic hill climbing algorithm that matches the
remarkable predictive accuracy of previous BART implementations, but is many
times faster and less memory intensive. Simulation studies show that the new
method is comparable in computation time and more accurate at function
estimation than both random forests and gradient boosting.
| stat |
Fast Convergence of Natural Gradient Descent for Overparameterized
Neural Networks | Natural gradient descent has proven effective at mitigating the effects of
pathological curvature in neural network optimization, but little is known
theoretically about its convergence properties, especially for \emph{nonlinear}
networks. In this work, we analyze for the first time the speed of convergence
of natural gradient descent on nonlinear neural networks with squared-error
loss. We identify two conditions which guarantee efficient convergence from
random initializations: (1) the Jacobian matrix (of network's output for all
training cases with respect to the parameters) has full row rank, and (2) the
Jacobian matrix is stable for small perturbations around the initialization.
For two-layer ReLU neural networks, we prove that these two conditions do in
fact hold throughout the training, under the assumptions of nondegenerate
inputs and overparameterization. We further extend our analysis to more general
loss functions. Lastly, we show that K-FAC, an approximate natural gradient
descent method, also converges to global minima under the same assumptions, and
we give a bound on the rate of this convergence.
| stat |
Towards a More Reliable Interpretation of Machine Learning Outputs for
Safety-Critical Systems using Feature Importance Fusion | When machine learning supports decision-making in safety-critical systems, it
is important to verify and understand the reasons why a particular output is
produced. Although feature importance calculation approaches assist in
interpretation, there is a lack of consensus regarding how features' importance
is quantified, which makes the explanations offered for the outcomes mostly
unreliable. A possible solution to address the lack of agreement is to combine
the results from multiple feature importance quantifiers to reduce the variance
of estimates. Our hypothesis is that this will lead to more robust and
trustworthy interpretations of the contribution of each feature to machine
learning predictions. To assist test this hypothesis, we propose an extensible
Framework divided in four main parts: (i) traditional data pre-processing and
preparation for predictive machine learning models; (ii) predictive machine
learning; (iii) feature importance quantification and (iv) feature importance
decision fusion using an ensemble strategy. We also introduce a novel fusion
metric and compare it to the state-of-the-art. Our approach is tested on
synthetic data, where the ground truth is known. We compare different fusion
approaches and their results for both training and test sets. We also
investigate how different characteristics within the datasets affect the
feature importance ensembles studied. Results show that our feature importance
ensemble Framework overall produces 15% less feature importance error compared
to existing methods. Additionally, results reveal that different levels of
noise in the datasets do not affect the feature importance ensembles' ability
to accurately quantify feature importance, whereas the feature importance
quantification error increases with the number of features and number of
orthogonal informative features.
| stat |
Non-Stationary Delayed Bandits with Intermediate Observations | Online recommender systems often face long delays in receiving feedback,
especially when optimizing for some long-term metrics. While mitigating the
effects of delays in learning is well-understood in stationary environments,
the problem becomes much more challenging when the environment changes. In
fact, if the timescale of the change is comparable to the delay, it is
impossible to learn about the environment, since the available observations are
already obsolete. However, the arising issues can be addressed if intermediate
signals are available without delay, such that given those signals, the
long-term behavior of the system is stationary. To model this situation, we
introduce the problem of stochastic, non-stationary, delayed bandits with
intermediate observations. We develop a computationally efficient algorithm
based on UCRL, and prove sublinear regret guarantees for its performance.
Experimental results demonstrate that our method is able to learn in
non-stationary delayed environments where existing methods fail.
| stat |
Error Estimation for Sketched SVD via the Bootstrap | In order to compute fast approximations to the singular value decompositions
(SVD) of very large matrices, randomized sketching algorithms have become a
leading approach. However, a key practical difficulty of sketching an SVD is
that the user does not know how far the sketched singular vectors/values are
from the exact ones. Indeed, the user may be forced to rely on analytical
worst-case error bounds, which do not account for the unique structure of a
given problem. As a result, the lack of tools for error estimation often leads
to much more computation than is really necessary. To overcome these
challenges, this paper develops a fully data-driven bootstrap method that
numerically estimates the actual error of sketched singular vectors/values. In
particular, this allows the user to inspect the quality of a rough initial
sketched SVD, and then adaptively predict how much extra work is needed to
reach a given error tolerance. Furthermore, the method is computationally
inexpensive, because it operates only on sketched objects, and it requires no
passes over the full matrix being factored. Lastly, the method is supported by
theoretical guarantees and a very encouraging set of experimental results.
| stat |
Is Temporal Difference Learning Optimal? An Instance-Dependent Analysis | We address the problem of policy evaluation in discounted Markov decision
processes, and provide instance-dependent guarantees on the $\ell_\infty$-error
under a generative model. We establish both asymptotic and non-asymptotic
versions of local minimax lower bounds for policy evaluation, thereby providing
an instance-dependent baseline by which to compare algorithms. Theory-inspired
simulations show that the widely-used temporal difference (TD) algorithm is
strictly suboptimal when evaluated in a non-asymptotic setting, even when
combined with Polyak-Ruppert iterate averaging. We remedy this issue by
introducing and analyzing variance-reduced forms of stochastic approximation,
showing that they achieve non-asymptotic, instance-dependent optimality up to
logarithmic factors.
| stat |
Sparse Gaussian Processes with Spherical Harmonic Features | We introduce a new class of inter-domain variational Gaussian processes (GP)
where data is mapped onto the unit hypersphere in order to use spherical
harmonic representations. Our inference scheme is comparable to variational
Fourier features, but it does not suffer from the curse of dimensionality, and
leads to diagonal covariance matrices between inducing variables. This enables
a speed-up in inference, because it bypasses the need to invert large
covariance matrices. Our experiments show that our model is able to fit a
regression model for a dataset with 6 million entries two orders of magnitude
faster compared to standard sparse GPs, while retaining state of the art
accuracy. We also demonstrate competitive performance on classification with
non-conjugate likelihoods.
| stat |
Bayesian Hierarchical Spatial Model for Small Area Estimation with
Non-ignorable Nonresponses and Its Applications to the NHANES Dental Caries
Assessments | The National Health and Nutrition Examination Survey (NHANES) is a major
program of the National Center for Health Statistics, designed to assess the
health and nutritional status of adults and children in the United States. The
analysis of NHANES dental caries data faces several challenges, including (1)
the data were collected using a complex, multistage, stratified,
unequal-probability sampling design; (2) the sample size of some primary
sampling units (PSU), e.g., counties, is very small; (3) the measures of dental
caries have complicated structure and correlation, and (4) there is a
substantial percentage of nonresponses, for which the missing data are expected
to be not missing at random or non-ignorable. We propose a Bayesian
hierarchical spatial model to address these analysis challenges. We develop a
two-level Potts model that closely resembles the caries evolution process and
captures complicated spatial correlations between teeth and surfaces of the
teeth. By adding Bayesian hierarchies to the Potts model, we account for the
multistage survey sampling design and also enable information borrowing across
PSUs for small area estimation. We incorporate sampling weights by including
them as a covariate in the model and adopt flexible B-splines to achieve robust
inference. We account for non-ignorable missing outcomes and covariates using
the selection model. We use data augmentation coupled with the noisy exchange
sampler to obtain the posterior of model parameters that involve
doubly-intractable normalizing constants. Our analysis results show strong
spatial associations between teeth and tooth surfaces and that dental hygienic
factors, fluorosis and sealant reduce the risks of having dental diseases.
| stat |
Two-phase analysis and study design for survival models with error-prone
exposures | Increasingly, medical research is dependent on data collected for
non-research purposes, such as electronic health records data (EHR). EHR data
and other large databases can be prone to measurement error in key exposures,
and unadjusted analyses of error-prone data can bias study results. Validating
a subset of records is a cost-effective way of gaining information on the error
structure, which in turn can be used to adjust analyses for this error and
improve inference. We extend the mean score method for the two-phase analysis
of discrete-time survival models, which uses the unvalidated covariates as
auxiliary variables that act as surrogates for the unobserved true exposures.
This method relies on a two-phase sampling design and an estimation approach
that preserves the consistency of complete case regression parameter estimates
in the validated subset, with increased precision leveraged from the auxiliary
data. Furthermore, we develop optimal sampling strategies which minimize the
variance of the mean score estimator for a target exposure under a fixed cost
constraint. We consider the setting where an internal pilot is necessary for
the optimal design so that the phase two sample is split into a pilot and an
adaptive optimal sample. Through simulations and data example, we evaluate
efficiency gains of the mean score estimator using the derived optimal
validation design compared to balanced and simple random sampling for the phase
two sample. We also empirically explore efficiency gains that the proposed
discrete optimal design can provide for the Cox proportional hazards model in
the setting of a continuous-time survival outcome.
| stat |
Statistical Properties of Exclusive and Non-exclusive Online Randomized
Experiments using Bucket Reuse | Randomized experiments is a key part of product development in the tech
industry. It is often necessary to run programs of exclusive experiments, i.e.,
experiments that cannot be run on the same units during the same time. These
programs implies restriction on the random sampling, as units that are
currently in an experiment cannot be sampled into a new one. Moreover, to
technically enable this type of coordination with large populations, the units
in the population are often grouped into 'buckets' and sampling is then
performed on the bucket level. This paper investigates some statistical
implications of both the restricted sampling and the bucket-level sampling. The
contribution of this paper is threefold: First, bucket sampling is connected to
the existing literature on randomized experiments in complex sampling designs
which enables establishing properties of the difference-in-means estimator of
the average treatment effect. These properties are needed for inference to the
population under random sampling of buckets. Second, the bias introduced by
restricting the sampling as imposed by programs of exclusive experiments, is
derived. Finally, simulation results supporting the theoretical findings are
presented together with recommendations on how to empirically evaluate and
handle this bias.
| stat |
Consistent Fixed-Effects Selection in Ultra-high dimensional Linear
Mixed Models with Error-Covariate Endogeneity | Recently, applied sciences, including longitudinal and clustered studies in
biomedicine require the analysis of ultra-high dimensional linear mixed effects
models where we need to select important fixed effect variables from a vast
pool of available candidates. However, all existing literature assume that all
the available covariates and random effect components are independent of the
model error which is often violated (endogeneity) in practice. In this paper,
we first investigate this important issue in ultra-high dimensional linear
mixed effects models with particular focus on the fixed effects selection. We
study the effects of different types of endogeneity on existing regularization
methods and prove their inconsistencies. Then, we propose a new profiled
focused generalized method of moments (PFGMM) approach to consistently select
fixed effects under 'error-covariate' endogeneity, i.e., in the presence of
correlation between the model error and covariates. Our proposal is proved to
be oracle consistent with probability tending to one and works well under most
other type of endogeneity too. Additionally, we also propose and illustrate a
few consistent parameter estimators, including those of the variance
components, along with variable selection through PFGMM. Empirical simulations
and an interesting real data example further support the claimed utility of our
proposal.
| stat |
Univariate Likelihood Projections and Characterizations of the
Multivariate Normal Distribution | The problem of characterizing a multivariate distribution of a random vector
using examination of univariate combinations of vector components is an
essential issue of multivariate analysis. The likelihood principle plays a
prominent role in developing powerful statistical inference tools. In this
context, we raise the question: can the univariate likelihood function based on
a random vector be used to provide the uniqueness in reconstructing the vector
distribution? In multivariate normal (MN) frameworks, this question links to a
reverse of Cochran's theorem that concerns the distribution of quadratic forms
in normal variables. We characterize the MN distribution through the univariate
likelihood type projections. The proposed principle is employed to illustrate
simple techniques for assessing multivariate normality via well-known tests
that use univariate observations. The presented testing strategy can exhibit
high and stable power characteristics in comparison to the well-known
procedures in various scenarios when observed vectors are non-MN distributed,
whereas their components are normally distributed random variables. In such
cases, the classical multivariate normality tests may break down completely.
KEY WORDS: Characterization, Goodness of fit, Infinity divisible, Likelihood,
Multivariate normal distribution, Projection, Quadratic form, Test for
multivariate normality.
| stat |
The Power of Batching in Multiple Hypothesis Testing | One important partition of algorithms for controlling the false discovery
rate (FDR) in multiple testing is into offline and online algorithms. The first
generally achieve significantly higher power of discovery, while the latter
allow making decisions sequentially as well as adaptively formulating
hypotheses based on past observations. Using existing methodology, it is
unclear how one could trade off the benefits of these two broad families of
algorithms, all the while preserving their formal FDR guarantees. To this end,
we introduce $\text{Batch}_{\text{BH}}$ and $\text{Batch}_{\text{St-BH}}$,
algorithms for controlling the FDR when a possibly infinite sequence of batches
of hypotheses is tested by repeated application of one of the most widely used
offline algorithms, the Benjamini-Hochberg (BH) method or Storey's improvement
of the BH method. We show that our algorithms interpolate between existing
online and offline methodology, thus trading off the best of both worlds.
| stat |
Learning-augmented count-min sketches via Bayesian nonparametrics | The count-min sketch (CMS) is a time and memory efficient randomized data
structure that provides estimates of tokens' frequencies in a data stream, i.e.
point queries, based on random hashed data. Learning-augmented CMSs improve the
CMS by learning models that allow to better exploit data properties. In this
paper, we focus on the learning-augmented CMS of Cai, Mitzenmacher and Adams
(\textit{NeurIPS} 2018), which relies on Bayesian nonparametric (BNP) modeling
of a data stream via Dirichlet process (DP) priors. This is referred to as the
CMS-DP, and it leads to BNP estimates of a point query as posterior means of
the point query given the hashed data. While BNPs is proved to be a powerful
tool for developing robust learning-augmented CMSs, ideas and methods behind
the CMS-DP are tailored to point queries under DP priors, and they can not be
used for other priors or more general queries. In this paper, we present an
alternative, and more flexible, derivation of the CMS-DP such that: i) it
allows to make use of the Pitman-Yor process (PYP) prior, which is arguably the
most popular generalization of the DP prior; ii) it can be readily applied to
the more general problem of estimating range queries. This leads to develop a
novel learning-augmented CMS under power-law data streams, referred to as the
CMS-PYP, which relies on BNP modeling of the stream via PYP priors.
Applications to synthetic and real data show that the CMS-PYP outperforms the
CMS and the CMS-DP in the estimation of low-frequency tokens; this known to be
a critical feature in natural language processing, where it is indeed common to
encounter power-law data streams.
| stat |
A neurally plausible model learns successor representations in partially
observable environments | Animals need to devise strategies to maximize returns while interacting with
their environment based on incoming noisy sensory observations. Task-relevant
states, such as the agent's location within an environment or the presence of a
predator, are often not directly observable but must be inferred using
available sensory information. Successor representations (SR) have been
proposed as a middle-ground between model-based and model-free reinforcement
learning strategies, allowing for fast value computation and rapid adaptation
to changes in the reward function or goal locations. Indeed, recent studies
suggest that features of neural responses are consistent with the SR framework.
However, it is not clear how such representations might be learned and computed
in partially observed, noisy environments. Here, we introduce a neurally
plausible model using distributional successor features, which builds on the
distributed distributional code for the representation and computation of
uncertainty, and which allows for efficient value function computation in
partially observed environments via the successor representation. We show that
distributional successor features can support reinforcement learning in noisy
environments in which direct learning of successful policies is infeasible.
| stat |
Learning pairwise Markov network structures using correlation
neighborhoods | Markov networks are widely studied and used throughout multivariate
statistics and computer science. In particular, the problem of learning the
structure of Markov networks from data without invoking chordality assumptions
in order to retain expressiveness of the model class has been given a
considerable attention in the recent literature, where numerous
constraint-based or score-based methods have been introduced. Here we develop a
new search algorithm for the network score-optimization that has several
computational advantages and scales well to high-dimensional data sets. The key
observation behind the algorithm is that the neighborhood of a variable can be
efficiently captured using local penalized likelihood ratio (PLR) tests by
exploiting an exponential decay of correlations across the neighborhood with an
increasing graph-theoretic distance from the focus node. The candidate
neighborhoods are then processed by a two-stage hill-climbing (HC) algorithm.
Our approach, termed fully as PLRHC-BIC$_{0.5}$, compares favorably against the
state-of-the-art methods in all our experiments spanning both low- and
high-dimensional networks and a wide range of sample sizes. An efficient
implementation of PLRHC-BIC$_{0.5}$ is freely available from the URL:
https://github.com/jurikuronen/plrhc.
| stat |
Predicting Pairwise Relations with Neural Similarity Encoders | Matrix factorization is at the heart of many machine learning algorithms, for
example, dimensionality reduction (e.g. kernel PCA) or recommender systems
relying on collaborative filtering. Understanding a singular value
decomposition (SVD) of a matrix as a neural network optimization problem
enables us to decompose large matrices efficiently while dealing naturally with
missing values in the given matrix. But most importantly, it allows us to learn
the connection between data points' feature vectors and the matrix containing
information about their pairwise relations. In this paper we introduce a novel
neural network architecture termed Similarity Encoder (SimEc), which is
designed to simultaneously factorize a given target matrix while also learning
the mapping to project the data points' feature vectors into a similarity
preserving embedding space. This makes it possible to, for example, easily
compute out-of-sample solutions for new data points. Additionally, we
demonstrate that SimEc can preserve non-metric similarities and even predict
multiple pairwise relations between data points at once.
| stat |
BETS: The dangers of selection bias in early analyses of the coronavirus
disease (COVID-19) pandemic | The coronavirus disease 2019 (COVID-19) has quickly grown from a regional
outbreak in Wuhan, China to a global pandemic. Early estimates of the epidemic
growth and incubation period of COVID-19 may have been biased due to sample
selection. Using detailed case reports from 14 locations in and outside
mainland China, we obtained 378 Wuhan-exported cases who left Wuhan before an
abrupt travel quarantine. We developed a generative model we call BETS for four
key epidemiological events---Beginning of exposure, End of exposure, time of
Transmission, and time of Symptom onset (BETS)---and derived explicit formulas
to correct for the sample selection. We gave a detailed illustration of why
some early and highly influential analyses of the COVID-19 pandemic were
severely biased. All our analyses, regardless of which subsample and model were
being used, point to an epidemic doubling time of 2 to 2.5 days during the
early outbreak in Wuhan. A Bayesian nonparametric analysis further suggests
that about 5% of the symptomatic cases may not develop symptoms within 14 days
of infection and that men may be much more likely than women to develop
symptoms within 2 days of infection.
| stat |
Improved scalability under heavy tails, without strong convexity | Real-world data is laden with outlying values. The challenge for machine
learning is that the learner typically has no prior knowledge of whether the
feedback it receives (losses, gradients, etc.) will be heavy-tailed or not. In
this work, we study a simple algorithmic strategy that can be leveraged when
both losses and gradients can be heavy-tailed. The core technique introduces a
simple robust validation sub-routine, which is used to boost the confidence of
inexpensive gradient-based sub-processes. Compared with recent robust gradient
descent methods from the literature, dimension dependence (both risk bounds and
cost) is substantially improved, without relying upon strong convexity or
expensive per-step robustification. Empirically, we also show that under
heavy-tailed losses, the proposed procedure cannot simply be replaced with
naive cross-validation. Taken together, we have a scalable method with
transparent guarantees, which performs well without prior knowledge of how
"convenient" the feedback it receives will be.
| stat |
Sparse Covariance Estimation in Logit Mixture Models | This paper introduces a new data-driven methodology for estimating sparse
covariance matrices of the random coefficients in logit mixture models.
Researchers typically specify covariance matrices in logit mixture models under
one of two extreme assumptions: either an unrestricted full covariance matrix
(allowing correlations between all random coefficients), or a restricted
diagonal matrix (allowing no correlations at all). Our objective is to find
optimal subsets of correlated coefficients for which we estimate covariances.
We propose a new estimator, called MISC, that uses a mixed-integer optimization
(MIO) program to find an optimal block diagonal structure specification for the
covariance matrix, corresponding to subsets of correlated coefficients, for any
desired sparsity level using Markov Chain Monte Carlo (MCMC) posterior draws
from the unrestricted full covariance matrix. The optimal sparsity level of the
covariance matrix is determined using out-of-sample validation. We demonstrate
the ability of MISC to correctly recover the true covariance structure from
synthetic data. In an empirical illustration using a stated preference survey
on modes of transportation, we use MISC to obtain a sparse covariance matrix
indicating how preferences for attributes are related to one another.
| stat |
Semiparametric Estimation of Cross-covariance Functions for Multivariate
Random Fields | The prevalence of spatially referenced multivariate data has impelled
researchers to develop a procedure for the joint modeling of multiple spatial
processes. This ordinarily involves modeling marginal and cross-process
dependence for any arbitrary pair of locations using a multivariate spatial
covariance function. However, building a flexible multivariate spatial
covariance function that is nonnegative definite is challenging. Here, we
propose a semiparametric approach for multivariate spatial covariance function
estimation with approximate Mat\'ern marginals and highly flexible
cross-covariance functions via their spectral representations. The flexibility
in our cross-covariance function arises due to B-spline based specification of
the underlying coherence functions, which in turn allows us to capture
non-trivial cross-spectral features. We then develop a likelihood-based
estimation procedure and perform multiple simulation studies to demonstrate the
performance of our method, especially on the coherence function estimation.
Finally, we analyze particulate matter concentrations ($\text{PM}_{2.5}$) and
wind speed data over the North-Eastern region of the United States, where we
illustrate that our proposed method outperforms the commonly used full
bivariate Mat\'ern model and the linear model of coregionalization for spatial
prediction.
| stat |
Selection of link function in binary regression: A case-study with world
happiness report on immigration | Selection of appropriate link function for binary regression remains an
important issue for data analysis and its influence on related inference. We
prescribe a new data-driven methodology to search for the same, considering
some popular classification assessment metrics. A case-study with World
Happiness report,2018 with special reference to immigration is presented for
demonstrating utility of the prescribed routine.
| stat |
Risk of the Least Squares Minimum Norm Estimator under the Spike
Covariance Model | We study risk of the minimum norm linear least squares estimator in when the
number of parameters $d$ depends on $n$, and $\frac{d}{n} \rightarrow \infty$.
We assume that data has an underlying low rank structure by restricting
ourselves to spike covariance matrices, where a fixed finite number of
eigenvalues grow with $n$ and are much larger than the rest of the eigenvalues,
which are (asymptotically) in the same order. We show that in this setting risk
of minimum norm least squares estimator vanishes in compare to risk of the null
estimator. We give asymptotic and non asymptotic upper bounds for this risk,
and also leverage the assumption of spike model to give an analysis of the bias
that leads to tighter bounds in compare to previous works.
| stat |
Measurement error in continuous endpoints in randomised trials: problems
and solutions | In randomised trials, continuous endpoints are often measured with some
degree of error. This study explores the impact of ignoring measurement error,
and proposes methods to improve statistical inference in the presence of
measurement error. Three main types of measurement error in continuous
endpoints are considered: classical, systematic and differential. For each
measurement error type, a corrected effect estimator is proposed. The corrected
estimators and several methods for confidence interval estimation are tested in
a simulation study. These methods combine information about error-prone and
error-free measurements of the endpoint in individuals not included in the
trial (external calibration sample). We show that if measurement error in
continuous endpoints is ignored, the treatment effect estimator is unbiased
when measurement error is classical, while Type-II error is increased at a
given sample size. Conversely, the estimator can be substantially biased when
measurement error is systematic or differential. In those cases, bias can
largely be prevented and inferences improved upon using information from an
external calibration sample, of which the required sample size increases as the
strength of the association between the error-prone and error-free endpoint
decreases. Measurement error correction using already a small (external)
calibration sample is shown to improve inferences and should be considered in
trials with error-prone endpoints. Implementation of the proposed correction
methods is accommodated by a new software package for R.
| stat |
Bayesian Approaches to Distribution Regression | Distribution regression has recently attracted much interest as a generic
solution to the problem of supervised learning where labels are available at
the group level, rather than at the individual level. Current approaches,
however, do not propagate the uncertainty in observations due to sampling
variability in the groups. This effectively assumes that small and large groups
are estimated equally well, and should have equal weight in the final
regression. We account for this uncertainty with a Bayesian distribution
regression formalism, improving the robustness and performance of the model
when group sizes vary. We frame our models in a neural network style, allowing
for simple MAP inference using backpropagation to learn the parameters, as well
as MCMC-based inference which can fully propagate uncertainty. We demonstrate
our approach on illustrative toy datasets, as well as on a challenging problem
of predicting age from images.
| stat |
Simulating longitudinal data from marginal structural models using the
additive hazard model | Observational longitudinal data on treatments and covariates are increasingly
used to investigate treatment effects, but are often subject to time-dependent
confounding. Marginal structural models (MSMs), estimated using inverse
probability of treatment weighting or the g-formula, are popular for handling
this problem. With increasing development of advanced causal inference methods,
it is important to be able to assess their performance in different scenarios
to guide their application. Simulation studies are a key tool for this, but
their use to evaluate causal inference methods has been limited. This paper
focuses on the use of simulations for evaluations involving MSMs in studies
with a time-to-event outcome. In a simulation, it is important to be able to
generate the data in such a way that the correct form of any models to be
fitted to those data is known. However, this is not straightforward in the
longitudinal setting because it is natural for data to be generated in a
sequential conditional manner, whereas MSMs involve fitting marginal rather
than conditional hazard models. We provide general results that enable the form
of the correctly-specified MSM to be derived based on a conditional data
generating procedure, and show how the results can be applied when the
conditional hazard model is an Aalen additive hazard or Cox model. Using
conditional additive hazard models is advantageous because they imply additive
MSMs that can be fitted using standard software. We describe and illustrate a
simulation algorithm. Our results will help researchers to effectively evaluate
causal inference methods via simulation.
| stat |
How to apply multiple imputation in propensity score matching with
partially observed confounders: a simulation study and practical
recommendations | Propensity score matching (PSM) has been widely used to mitigate confounding
in observational studies, although complications arise when the covariates used
to estimate the PS are only partially observed. Multiple imputation (MI) is a
potential solution for handling missing covariates in the estimation of the PS.
Unfortunately, it is not clear how to best apply MI strategies in the context
of PSM. We conducted a simulation study to compare the performances of popular
non-MI missing data methods and various MI-based strategies under different
missing data mechanisms (MDMs). We found that commonly applied missing data
methods resulted in biased and inefficient estimates, and we observed large
variation in performance across MI-based strategies. Based on our findings, we
recommend 1) deriving the PS after applying MI (referred to as MI-derPassive);
2) conducting PSM within each imputed data set followed by averaging the
treatment effects to arrive at one summarized finding (INT-within) for mild
MDMs and averaging the PSs across multiply imputed datasets before obtaining
one treatment effect using PSM (INT-across) for more complex MDMs; 3) a
bootstrapped-based variance to account for uncertainty of PS estimation,
matching, and imputation; and 4) inclusion of key auxiliary variables in the
imputation model.
| stat |
Comments on Leo Breiman's paper 'Statistical Modeling: The Two Cultures'
(Statistical Science, 2001, 16(3), 199-231) | Breiman challenged statisticians to think more broadly, to step into the
unknown, model-free learning world, with him paving the way forward. Statistics
community responded with slight optimism, some skepticism, and plenty of
disbelief. Today, we are at the same crossroad anew. Faced with the enormous
practical success of model-free, deep, and machine learning, we are naturally
inclined to think that everything is resolved. A new frontier has emerged; the
one where the role, impact, or stability of the {\it learning} algorithms is no
longer measured by prediction quality, but an inferential one -- asking the
questions of {\it why} and {\it if} can no longer be safely ignored.
| stat |
(Decision and regression) tree ensemble based kernels for regression and
classification | Tree based ensembles such as Breiman's random forest (RF) and Gradient
Boosted Trees (GBT) can be interpreted as implicit kernel generators, where the
ensuing proximity matrix represents the data-driven tree ensemble kernel.
Kernel perspective on the RF has been used to develop a principled framework
for theoretical investigation of its statistical properties. Recently, it has
been shown that the kernel interpretation is germane to other tree-based
ensembles e.g. GBTs. However, practical utility of the links between kernels
and the tree ensembles has not been widely explored and systematically
evaluated.
Focus of our work is investigation of the interplay between kernel methods
and the tree based ensembles including the RF and GBT. We elucidate the
performance and properties of the RF and GBT based kernels in a comprehensive
simulation study comprising of continuous and binary targets. We show that for
continuous targets, the RF/GBT kernels are competitive to their respective
ensembles in higher dimensional scenarios, particularly in cases with larger
number of noisy features. For the binary target, the RF/GBT kernels and their
respective ensembles exhibit comparable performance. We provide the results
from real life data sets for regression and classification to show how these
insights may be leveraged in practice. Overall, our results support the tree
ensemble based kernels as a valuable addition to the practitioner's toolbox.
Finally, we discuss extensions of the tree ensemble based kernels for
survival targets, interpretable prototype and landmarking classification and
regression. We outline future line of research for kernels furnished by
Bayesian counterparts of the frequentist tree ensembles.
| stat |
Set-valued classification -- overview via a unified framework | Multi-class classification problem is among the most popular and well-studied
statistical frameworks. Modern multi-class datasets can be extremely ambiguous
and single-output predictions fail to deliver satisfactory performance. By
allowing predictors to predict a set of label candidates, set-valued
classification offers a natural way to deal with this ambiguity. Several
formulations of set-valued classification are available in the literature and
each of them leads to different prediction strategies. The present survey aims
to review popular formulations using a unified statistical framework. The
proposed framework encompasses previously considered and leads to new
formulations as well as it allows to understand underlying trade-offs of each
formulation. We provide infinite sample optimal set-valued classification
strategies and review a general plug-in principle to construct data-driven
algorithms. The exposition is supported by examples and pointers to both
theoretical and practical contributions. Finally, we provide experiments on
real-world datasets comparing these approaches in practice and providing
general practical guidelines.
| stat |
Sampling Humans for Optimizing Preferences in Coloring Artwork | Many circumstances of practical importance have performance or success
metrics which exist implicitly---in the eye of the beholder, so to speak.
Tuning aspects of such problems requires working without defined metrics and
only considering pairwise comparisons or rankings. In this paper, we review an
existing Bayesian optimization strategy for determining most-preferred
outcomes, and identify an adaptation to allow it to handle ties. We then
discuss some of the issues we have encountered when humans use this
optimization strategy to optimize coloring a piece of abstract artwork. We hope
that, by participating in this workshop, we can learn how other researchers
encounter difficulties unique to working with humans in the loop.
| stat |
Active learning for binary classification with variable selection | Modern computing and communication technologies can make data collection
procedures very efficient. However, our ability to analyze large data sets
and/or to extract information out from them is hard-pressed to keep up with our
capacities for data collection. Among these huge data sets, some of them are
not collected for any particular research purpose. For a classification
problem, this means that the essential label information may not be readily
obtainable, in the data set in hands, and an extra labeling procedure is
required such that we can have enough label information to be used for
constructing a classification model. When the size of a data set is huge, to
label each subject in it will cost a lot in both capital and time. Thus, it is
an important issue to decide which subjects should be labeled first in order to
efficiently reduce the training cost/time. Active learning method is a
promising outlet for this situation, because with the active learning ideas, we
can select the unlabeled subjects sequentially without knowing their label
information. In addition, there will be no confirmed information about the
essential variables for constructing an efficient classification rule. Thus,
how to merge a variable selection scheme with an active learning procedure is
of interest. In this paper, we propose a procedure for building binary
classification models when the complete label information is not available in
the beginning of the training stage. We study an model-based active learning
procedure with sequential variable selection schemes, and discuss the results
of the proposed procedure from both theoretical and numerical aspects.
| stat |
The Scenario Culture | Scenario Analysis is a risk assessment tool that aims to evaluate the impact
of a small number of distinct plausible future scenarios. In this paper, we
provide an overview of important aspects of Scenario Analysis including when it
is appropriate, the design of scenarios, uncertainty and encouraging
creativity. Each of these issues is discussed in the context of climate, energy
and legal scenarios.
| stat |
Spherical Regression under Mismatch Corruption with Application to
Automated Knowledge Translation | Motivated by a series of applications in data integration, language
translation, bioinformatics, and computer vision, we consider spherical
regression with two sets of unit-length vectors when the data are corrupted by
a small fraction of mismatch in the response-predictor pairs. We propose a
three-step algorithm in which we initialize the parameters by solving an
orthogonal Procrustes problem to estimate a translation matrix $\mathbb{W}$
ignoring the mismatch. We then estimate a mapping matrix aiming to correct the
mismatch using hard-thresholding to induce sparsity, while incorporating
potential group information. We eventually obtain a refined estimate for
$\mathbb{W}$ by removing the estimated mismatched pairs. We derive the error
bound for the initial estimate of $\mathbb{W}$ in both fixed and
high-dimensional setting. We demonstrate that the refined estimate of
$\mathbb{W}$ achieves an error rate that is as good as if no mismatch is
present. We show that our mapping recovery method not only correctly
distinguishes one-to-one and one-to-many correspondences, but also consistently
identifies the matched pairs and estimates the weight vector for combined
correspondence. We examine the finite sample performance of the proposed method
via extensive simulation studies, and with application to the unsupervised
translation of medical codes using electronic health records data.
| stat |
Online Structural Change-point Detection of High-dimensional Streaming
Data via Dynamic Sparse Subspace Learning | High-dimensional streaming data are becoming increasingly ubiquitous in many
fields. They often lie in multiple low-dimensional subspaces, and the manifold
structures may change abruptly on the time scale due to pattern shift or
occurrence of anomalies. However, the problem of detecting the structural
changes in a real-time manner has not been well studied. To fill this gap, we
propose a dynamic sparse subspace learning (DSSL) approach for online
structural change-point detection of high-dimensional streaming data. A novel
multiple structural change-point model is proposed and it is shown to be
equivalent to maximizing a posterior under certain conditions. The asymptotic
properties of the estimators are investigated. The penalty coefficients in our
model can be selected by AMDL criterion based on some historical data. An
efficient Pruned Exact Linear Time (PELT) based method is proposed for online
optimization and change-point detection. The effectiveness of the proposed
method is demonstrated through a simulation study and a real case study using
gesture data for motion tracking.
| stat |
Lipschitz neural networks are dense in the set of all Lipschitz
functions | This note shows that, for a fixed Lipschitz constant $L > 0$, one layer
neural networks that are $L$-Lipschitz are dense in the set of all
$L$-Lipschitz functions with respect to the uniform norm on bounded sets.
| stat |
ISLET: Fast and Optimal Low-rank Tensor Regression via Importance
Sketching | In this paper, we develop a novel procedure for low-rank tensor regression,
namely \emph{\underline{I}mportance \underline{S}ketching \underline{L}ow-rank
\underline{E}stimation for \underline{T}ensors} (ISLET). The central idea
behind ISLET is \emph{importance sketching}, i.e., carefully designed sketches
based on both the responses and low-dimensional structure of the parameter of
interest. We show that the proposed method is sharply minimax optimal in terms
of the mean-squared error under low-rank Tucker assumptions and under
randomized Gaussian ensemble design. In addition, if a tensor is low-rank with
group sparsity, our procedure also achieves minimax optimality. Further, we
show through numerical study that ISLET achieves comparable or better
mean-squared error performance to existing state-of-the-art methods while
having substantial storage and run-time advantages including capabilities for
parallel and distributed computing. In particular, our procedure performs
reliable estimation with tensors of dimension $p = O(10^8)$ and is $1$ or $2$
orders of magnitude faster than baseline methods.
| stat |
Rapid adjustment and post-processing of temperature forecast
trajectories | Modern weather forecasts are commonly issued as consistent multi-day forecast
trajectories with a time resolution of 1-3 hours. Prior to issuing, statistical
post-processing is routinely used to correct systematic errors and
misrepresentations of the forecast uncertainty. However, once the forecast has
been issued, it is rarely updated before it is replaced in the next forecast
cycle of the numerical weather prediction (NWP) model. This paper shows that
the error correlation structure within the forecast trajectory can be utilized
to substantially improve the forecast between the NWP forecast cycles by
applying additional post-processing steps each time new observations become
available. The proposed rapid adjustment is applied to temperature forecast
trajectories from the UK Met Office's convective-scale ensemble MOGREPS-UK.
MOGREPS-UK is run four times daily and produces hourly forecasts for up to 36
hours ahead. Our results indicate that the rapidly adjusted forecast from the
previous NWP forecast cycle outperforms the new forecast for the first few
hours of the next cycle, or until the new forecast itself can be rapidly
adjusted, suggesting a new strategy for updating the forecast cycle.
| stat |
An Improved and Extended Bayesian Synthetic Control | An improved and extended Bayesian synthetic control model is presented,
expanding upon the latent factor model in Tuomaala 2019. The changes we make
include 1) standardization of the data prior to model fit - which improves
efficiency and generalization across different data sets; 2) adding time
varying covariates; 3) adding the ability to have multiple treated units; 4)
fitting the latent factors within the Bayesian model; and, 5) a sparsity
inducing prior to automatically tune the number of latent factors. We
demonstrate the similarity of estimates to two traditional synthetic control
studies in Abadie, Diamond, and Hainmueller 2010 and Abadie, Diamond, and
Hainmueller 2015 and extend to multiple target series with a new example of
estimating digital website visitation from changes in data collection due to
digital privacy laws.
| stat |
Estimating Individual Treatment Effects using Non-Parametric Regression
Models: a Review | Large observational data are increasingly available in disciplines such as
health, economic and social sciences, where researchers are interested in
causal questions rather than prediction. In this paper, we investigate the
problem of estimating heterogeneous treatment effects using non-parametric
regression-based methods. Firstly, we introduce the setup and the issues
related to conducting causal inference with observational or non-fully
randomized data, and how these issues can be tackled with the help of
statistical learning tools. Then, we provide a review of state-of-the-art
methods, with a particular focus on non-parametric modeling, and we cast them
under a unifying taxonomy. After presenting a brief overview on the problem of
model selection, we illustrate the performance of some of the methods on three
different simulated studies and on a real world example to investigate the
effect of participation in school meal programs on health indicators.
| stat |
Factor Models for High-Dimensional Tensor Time Series | Large tensor (multi-dimensional array) data are now routinely collected in a
wide range of applications, due to modern data collection capabilities. Often
such observations are taken over time, forming tensor time series. In this
paper we present a factor model approach for analyzing high-dimensional dynamic
tensor time series and multi-category dynamic transport networks. Two
estimation procedures along with their theoretical properties and simulation
results are presented. Two applications are used to illustrate the model and
its interpretations.
| stat |
Statistics students' identification of inferential model elements within
contexts of their own invention | Statistical thinking partially depends upon an iterative process by which
essential features of a problem setting are identified and mapped onto an
abstract model or archetype, and then translated back into the context of the
original problem setting (Wild and Pfannkuch 1999). Assessment in introductory
statistics often relies on tasks that present students with data in context and
expects them to choose and describe an appropriate model. This study explores
post-secondary student responses to an alternative task that prompts students
to clearly identify a sample, population, statistic, and parameter using a
context of their own invention. The data include free text narrative responses
of a random sample of 500 students from a sample of more than 1600 introductory
statistics students. Results suggest that students' responses often portrayed
sample and population accurately. Portrayals of statistic and parameter were
less reliable and were associated with descriptions of a wide variety of other
concepts. Responses frequently attributed a variable of some kind to the
statistic, or a study design detail to the parameter. Implications for
instruction and research are discussed, including a call for emphasis on a
modeling paradigm in introductory statistics.
| stat |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.