title
stringlengths 6
244
| abstract
stringlengths 19
5.09k
| label
stringclasses 10
values |
|---|---|---|
Permutation-Based True Discovery Guarantee by Sum Tests | Sum-based global tests are highly popular in multiple hypothesis testing. In
this paper we propose a general closed testing procedure for sum tests, which
provides lower confidence bounds for the proportion of true discoveries (TDP),
simultaneously over all subsets of hypotheses; these simultaneous inferences
come for free, i.e., without any adjustment of the alpha-level, whenever a
global test is used. Our method allows for an exploratory approach, as
simultaneity ensures control of the TDP even when the subset of interest is
selected post hoc. It adapts to the unknown joint distribution of the data
through permutation testing. Any sum test may be employed, depending on the
desired power properties. We present an iterative shortcut for the closed
testing procedure, based on the branch and bound algorithm, which converges to
the full closed testing results, often after few iterations; even if it is
stopped early, it controls the TDP. The feasibility of the method for high
dimensional data is illustrated on brain imaging data, then we compare the
properties of different choices for the sum test through simulations.
| stat |
Nonmyopic and pseudo-nonmyopic approaches to optimal sequential design
in the presence of covariates | In sequential experiments, subjects become available for the study over a
period of time, and covariates are often measured at the time of arrival. We
consider the setting where the sample size is fixed but covariate values are
unknown until subjects enrol. Given a model for the outcome, a sequential
optimal design approach can be used to allocate treatments to minimize the
variance of the treatment effect. We extend existing optimal design methodology
so it can be used within a nonmyopic framework, where treatment allocation for
the current subject depends not only on the treatments and covariates of the
subjects already enrolled in the study, but also the impact of possible future
treatment assignments. The nonmyopic approach is computationally expensive as
it requires recursive formulae. We propose a pseudo-nonmyopic approach which
has a similar aim to the nonmyopic approach, but does not involve recursion and
instead relies on simulations of future possible decisions. Our simulation
studies show that the myopic approach is the most efficient for the logistic
model case with a single binary covariate and binary treatment.
| stat |
Interpretable Sparse Proximate Factors for Large Dimensions | This paper proposes sparse and easy-to-interpret proximate factors to
approximate statistical latent factors. Latent factors in a large-dimensional
factor model can be estimated by principal component analysis (PCA), but are
usually hard to interpret. We obtain proximate factors that are easier to
interpret by shrinking the PCA factor weights and setting them to zero except
for the largest absolute ones. We show that proximate factors constructed with
only 5-10% of the data are usually sufficient to almost perfectly replicate the
population and PCA factors without actually assuming a sparse structure in the
weights or loadings. Using extreme value theory we explain why sparse proximate
factors can be substitutes for non-sparse PCA factors. We derive analytical
asymptotic bounds for the correlation of appropriately rotated proximate
factors with the population factors. These bounds provide guidance on how to
construct the proximate factors. In simulations and empirical analyses of
financial portfolio and macroeconomic data we illustrate that sparse proximate
factors are close substitutes for PCA factors with average correlations of
around 97.5% while being interpretable.
| stat |
Ensemble Riemannian Data Assimilation over the Wasserstein Space | In this paper, we present an ensemble data assimilation paradigm over a
Riemannian manifold equipped with the Wasserstein metric. Unlike the Eulerian
penalization of error in the Euclidean space, the Wasserstein metric can
capture translation and difference between the shapes of square-integrable
probability distributions of the background state and observations -- enabling
to formally penalize geophysical biases in state-space with non-Gaussian
distributions. The new approach is applied to dissipative and chaotic
evolutionary dynamics and its potential advantages and limitations are
highlighted compared to the classic variational and filtering data assimilation
approaches under systematic and random errors.
| stat |
Machine Learning in Downlink Coordinated Multipoint in Heterogeneous
Networks | We propose a method for downlink coordinated multipoint (DL CoMP) in
heterogeneous fifth generation New Radio (NR) networks. The primary
contribution of our paper is an algorithm to enhance the trigger of DL CoMP
using online machine learning. We use support vector machine (SVM) classifiers
to enhance the user downlink throughput in a realistic frequency division
duplex network environment. Our simulation results show improvement in both the
macro and pico base station downlink throughputs due to the informed triggering
of the multiple radio streams as learned by the SVM classifier.
| stat |
Unified Robust Estimation via the COCO | Robust estimation is concerned with how to provide reliable parameter
estimates in the presence of outliers. Numerous robust loss functions have been
proposed in regression and classification, along with various computing
algorithms. This article proposes a unified framework for loss function
construction and parameter estimation. The CC-family contains composite of
concave and convex functions. The properties of the CC-family are investigated,
and CC-estimation is innovatively conducted via composite optimization by
conjugation operator (COCO). The weighted estimators are simple to implement,
demonstrate robust quality in penalized generalized linear models and support
vector machines, and can be conveniently extended to even more broad
applications with existing software.
| stat |
nprobust: Nonparametric Kernel-Based Estimation and Robust
Bias-Corrected Inference | Nonparametric kernel density and local polynomial regression estimators are
very popular in Statistics, Economics, and many other disciplines. They are
routinely employed in applied work, either as part of the main empirical
analysis or as a preliminary ingredient entering some other estimation or
inference procedure. This article describes the main methodological and
numerical features of the software package nprobust, which offers an array of
estimation and inference procedures for nonparametric kernel-based density and
local polynomial regression methods, implemented in both the R and Stata
statistical platforms. The package includes not only classical bandwidth
selection, estimation, and inference methods (Wand and Jones, 1995; Fan and
Gijbels, 1996), but also other recent developments in the statistics and
econometrics literatures such as robust bias-corrected inference and coverage
error optimal bandwidth selection (Calonico, Cattaneo and Farrell, 2018, 2019).
Furthermore, this article also proposes a simple way of estimating optimal
bandwidths in practice that always delivers the optimal mean square error
convergence rate regardless of the specific evaluation point, that is, no
matter whether it is implemented at a boundary or interior point. Numerical
performance is illustrated using an empirical application and simulated data,
where a detailed numerical comparison with other R packages is given.
| stat |
Smooth function approximation by deep neural networks with general
activation functions | There has been a growing interest in expressivity of deep neural networks.
However, most of the existing work about this topic focuses only on the
specific activation function such as ReLU or sigmoid. In this paper, we
investigate the approximation ability of deep neural networks with a broad
class of activation functions. This class of activation functions includes most
of frequently used activation functions. We derive the required depth, width
and sparsity of a deep neural network to approximate any H\"older smooth
function upto a given approximation error for the large class of activation
functions. Based on our approximation error analysis, we derive the minimax
optimality of the deep neural network estimators with the general activation
functions in both regression and classification problems.
| stat |
Exact computation of projection regression depth and fast computation of
its induced median and other estimators | Zuo (2019) (Z19) addressed the computation of the projection regression depth
(PRD) and its induced median (the maximum depth estimator). Z19 achieved the
exact computation of PRD via a modified version of regular univariate sample
median, which resulted in the loss of invariance of PRD and the equivariance of
depth induced median. This article achieves the exact computation without
scarifying the invariance of PRD and the equivariance of the regression median.
Z19 also addressed the approximate computation of PRD induced median, the naive
algorithm in Z19 is very slow. This article modifies the approximation in Z19
and adopts Rcpp package and consequently obtains a much (could be $100$ times)
faster algorithm with an even better level of accuracy meanwhile. Furthermore,
as the third major contribution, this article introduces three new depth
induced estimators which can run $300$ times faster than that of Z19 meanwhile
maintaining the same (or a bit better) level of accuracy. Real as well as
simulated data examples are presented to illustrate the difference between the
algorithms of Z19 and the ones proposed in this article. Findings support the
statements above and manifest the major contributions of the article.
| stat |
Integer-valued autoregressive process with flexible marginal and
innovation distributions | INteger Auto-Regressive (INAR) processes are usually defined by specifying
the innovations and the operator, which often leads to difficulties in deriving
marginal properties of the process. In many practical situations, a major
modeling limitation is that it is difficult to justify the choice of the
operator. To overcome these drawbacks, we propose a new flexible approach to
build an INAR model: we pre-specify the marginal and innovation distributions.
Hence, the operator is a consequence of specifying the desired marginal and
innovation distributions. Our new INAR model has both marginal and innovations
geometric distributed, being a direct alternative to the classical Poisson INAR
model. Our proposed process has interesting stochastic properties such as an
MA($\infty$) representation, time-reversibility, and closed-forms for the
transition probabilities $h$-steps ahead, allowing for coherent forecasting. We
analyze time-series counts of skin lesions using our proposed approach,
comparing it with existing INAR and INGARCH models. Our model gives more
adherence to the data and better forecasting performance.
| stat |
To update or not to update? Delayed Nonparametric Bandits with
Randomized Allocation | Delayed rewards problem in contextual bandits has been of interest in various
practical settings. We study randomized allocation strategies and provide an
understanding on how the exploration-exploitation tradeoff is affected by
delays in observing the rewards. In randomized strategies, the extent of
exploration-exploitation is controlled by a user-determined exploration
probability sequence. In the presence of delayed rewards, one may choose
between using the original exploration sequence that updates at every time
point or update the sequence only when a new reward is observed, leading to two
competing strategies. In this work, we show that while both strategies may lead
to strong consistency in allocation, the property holds for a wider scope of
situations for the latter. However, for finite sample performance, we
illustrate that both strategies have their own advantages and disadvantages,
depending on the severity of the delay and underlying reward generating
mechanisms.
| stat |
Bracketing in the Comparative Interrupted Time-Series Design to Address
Concerns about History Interacting with Group: Evaluating Missouri Handgun
Purchaser Law | In the comparative interrupted time series design (also called the method of
difference-in-differences), the change in outcome in a group exposed to
treatment in the periods before and after the exposure is compared to the
change in outcome in a control group not exposed to treatment in either period.
The standard difference-in-difference estimator for a comparative interrupted
time series design will be biased for estimating the causal effect of the
treatment if there is an interaction between history in the after period and
the groups; for example, there is a historical event besides the start of the
treatment in the after period that benefits the treated group more than the
control group. We present a bracketing method for bounding the effect of an
interaction between history and the groups that arises from a time-invariant
unmeasured confounder having a different effect in the after period than the
before period. The method is applied to a study of the effect of the repeal of
Missouri's permit-to-purchase handgun law on its firearm homicide rate. We
estimate that the effect of the permit-to-purchase repeal on Missouri's firearm
homicide rate is bracketed between 0.9 and 1.3 homicides per 100,000 people,
corresponding to a percentage increase of 17% to 27% (95% confidence interval:
[0.6,1.7] or [11%,35%]). A placebo study provides additional support for the
hypothesis that the repeal has a causal effect of increasing the rate of
state-wide firearm homicides.
| stat |
Benchmarking survival outcomes: A funnel plot for survival data | Benchmarking is commonly used in many healthcare settings to monitor clinical
performance, with the aim of increasing cost-effectiveness and safe care of
patients. The funnel plot is a popular tool in visualizing the performance of a
healthcare center in relation to other centers and to a target, taking into
account statistical uncertainty. In this paper we develop methodology for
constructing funnel plots for survival data. The method takes into account
censoring and can deal with differences in censoring distributions across
centers. Practical issues in implementing the methodology are discussed,
particularly in the setting of benchmarking clinical outcomes for hematopoietic
stem cell transplantation. A simulation study is performed to assess the
performance of the funnel plots under several scenarios. Our methodology is
illustrated using data from the EBMT benchmarking project.
| stat |
Hollow-tree Super: a directional and scalable approach for feature
importance in boosted tree models | Current limitations in boosted tree modelling prevent the effective scaling
to datasets with a large feature number, particularly when investigating the
magnitude and directionality of various features on classification. We present
a novel methodology, Hollow-tree Super (HOTS), to resolve and visualize feature
importance in boosted tree models involving a large number of features.
Further, HOTS allows for investigation of the directionality and magnitude
various features have on classification. Using the Iris dataset, we first
compare HOTS to Gini Importance, Partial Dependence Plots, and Permutation
Importance, and demonstrate how HOTS resolves the weaknesses present in these
methods. We then show how HOTS can be utilized in high dimensional
neuroscientific data, by taking 60 Schizophrenic subjects and applying the
method to determine which brain regions were most important for classification
of schizophrenia as determined by the PANSS. HOTS effectively replicated and
supported the findings of Gini importance, Partial Dependence Plots and
Permutation importance within the Iris dataset. When applied to the
schizophrenic brain dataset, HOTS was able to resolve the top 10 most important
features for classification, as well as their directionality for classification
and magnitude compared to other features. Cross-validation supported that these
same 10 features were consistently used in the decision-making process across
multiple trees, and these features were localised primarily to the occipital
and parietal cortices, commonly disturbed brain regions in those with
Schizophrenia. It is imperative that a methodology is developed that is able to
handle the demands of working with large datasets that contain a large number
of features. HOTS represents a unique way to investigate both the
directionality and magnitude of feature importance when working at scale with
boosted-tree modelling.
| stat |
Online Learning to Rank with Features | We introduce a new model for online ranking in which the click probability
factors into an examination and attractiveness function and the attractiveness
function is a linear function of a feature vector and an unknown parameter.
Only relatively mild assumptions are made on the examination function. A novel
algorithm for this setup is analysed, showing that the dependence on the number
of items is replaced by a dependence on the dimension, allowing the new
algorithm to handle a large number of items. When reduced to the orthogonal
case, the regret of the algorithm improves on the state-of-the-art.
| stat |
Discovering Diverse Solutions in Deep Reinforcement Learning | Reinforcement learning (RL) algorithms are typically limited to learning a
single solution of a specified task, even though there often exists diverse
solutions to a given task. Compared with learning a single solution, learning a
set of diverse solutions is beneficial because diverse solutions enable robust
few-shot adaptation and allow the user to select a preferred solution. Although
previous studies have showed that diverse behaviors can be modeled with a
policy conditioned on latent variables, an approach for modeling an infinite
set of diverse solutions with continuous latent variables has not been
investigated. In this study, we propose an RL method that can learn infinitely
many solutions by training a policy conditioned on a continuous or discrete
low-dimensional latent variable. Through continuous control tasks, we
demonstrate that our method can learn diverse solutions in a data-efficient
manner and that the solutions can be used for few-shot adaptation to solve
unseen tasks.
| stat |
Joint integrative analysis of multiple data sources with correlated
vector outcomes | We propose a distributed quadratic inference function framework to jointly
estimate regression parameters from multiple potentially heterogeneous data
sources with correlated vector outcomes. The primary goal of this joint
integrative analysis is to estimate covariate effects on all outcomes through a
marginal regression model in a statistically and computationally efficient way.
We develop a data integration procedure for statistical estimation and
inference of regression parameters that is implemented in a fully distributed
and parallelized computational scheme. To overcome computational and modeling
challenges arising from the high-dimensional likelihood of the correlated
vector outcomes, we propose to analyze each data source using Qu, Lindsay and
Li (2000)'s quadratic inference functions, and then to jointly reestimate
parameters from each data source by accounting for correlation between data
sources using a combined meta-estimator in a similar spirit to Hansen (1982)'s
generalised method of moments. We show both theoretically and numerically that
the proposed method yields efficiency improvements and is computationally fast.
We illustrate the proposed methodology with the joint integrative analysis of
the association between smoking and metabolites in a large multi-cohort study
and provide an R package for ease of implementation.
| stat |
Robust Nonparametric Difference-in-Differences Estimation | We consider the problem of treatment effect estimation in
difference-in-differences designs where parallel trends hold only after
conditioning on covariates. Existing methods for this problem rely on strong
additional assumptions, e.g., that any covariates may only have linear effects
on the outcome of interest, or that there is no covariate shift between
different cross sections taken in the same state. Here, we develop a suite of
new methods for nonparametric difference-in-differences estimation that require
essentially no assumptions beyond conditional parallel trends and a relevant
form of overlap. Our proposals show promising empirical performance across a
variety of simulation setups, and are more robust than the standard methods
based on either linear regression or propensity weighting.
| stat |
Continuum-Armed Bandits: A Function Space Perspective | Continuum-armed bandits (a.k.a., black-box or $0^{th}$-order optimization)
involves optimizing an unknown objective function given an oracle that
evaluates the function at a query point, with the goal of using as few query
points as possible. In the most well-studied case, the objective function is
assumed to be Lipschitz continuous and minimax rates of simple and cumulative
regrets are known in both noiseless and noisy settings. This paper studies
continuum-armed bandits under more general smoothness conditions, namely Besov
smoothness conditions, on the objective function. In both noiseless and noisy
conditions, we derive minimax rates under simple and cumulative regrets. Our
results show that minimax rates over objective functions in a Besov space are
identical to minimax rates over objective functions in the smallest H\"older
space into which the Besov space embeds.
| stat |
Rank aggregation for non-stationary data streams | We consider the problem of learning over non-stationary ranking streams. The
rankings can be interpreted as the preferences of a population and the
non-stationarity means that the distribution of preferences changes over time.
Our goal is to learn, in an online manner, the current distribution of
rankings. The bottleneck of this process is a rank aggregation problem.
We propose a generalization of the Borda algorithm for non-stationary ranking
streams. Moreover, we give bounds on the minimum number of samples required to
output the ground truth with high probability. Besides, we show how the optimal
parameters are set. Then, we generalize the whole family of weighted voting
rules (the family to which Borda belongs) to situations in which some rankings
are more \textit{reliable} than others and show that this generalization can
solve the problem of rank aggregation over non-stationary data streams.
| stat |
Robust High Dimensional Expectation Maximization Algorithm via Trimmed
Hard Thresholding | In this paper, we study the problem of estimating latent variable models with
arbitrarily corrupted samples in high dimensional space ({\em i.e.,} $d\gg n$)
where the underlying parameter is assumed to be sparse. Specifically, we
propose a method called Trimmed (Gradient) Expectation Maximization which adds
a trimming gradients step and a hard thresholding step to the Expectation step
(E-step) and the Maximization step (M-step), respectively. We show that under
some mild assumptions and with an appropriate initialization, the algorithm is
corruption-proofing and converges to the (near) optimal statistical rate
geometrically when the fraction of the corrupted samples $\epsilon$ is bounded
by $ \tilde{O}(\frac{1}{\sqrt{n}})$. Moreover, we apply our general framework
to three canonical models: mixture of Gaussians, mixture of regressions and
linear regression with missing covariates. Our theory is supported by thorough
numerical results.
| stat |
The Debiased Spatial Whittle Likelihood | We provide a computationally and statistically efficient method for
estimating the parameters of a stochastic Gaussian model observed on a regular
spatial grid in any number of dimensions. Our proposed method, which we call
the debiased spatial Whittle likelihood, makes important corrections to the
well-known Whittle likelihood to account for large sources of bias caused by
boundary effects and aliasing. We generalise the approach to flexibly allow for
significant volumes of missing data, for the usage of irregular sampling
schemes including those with lower-dimensional substructure, and for irregular
sampling boundaries. We build a theoretical framework under relatively weak
assumptions which ensures consistency and asymptotic normality in numerous
practical settings. We provide detailed implementation guidelines which ensure
the estimation procedure can still be conducted in $\mathcal{O}(n\log n)$
operations, where $n$ is the number of points of the encapsulating rectangular
grid, thus keeping the computational scalability of Fourier and Whittle-based
methods for large data sets. We validate our procedure over a range of
simulated and real world settings, and compare with state-of-the-art
alternatives, demonstrating the enduring significant practical appeal of
Fourier-based methods, provided they are corrected by the constructive
procedures developed in this paper.
| stat |
Tractable Bayesian Density Regression via Logit Stick-Breaking Priors | There is a growing interest in learning how the distribution of a response
variable changes with a set of predictors. Bayesian nonparametric dependent
mixture models provide a flexible approach to address this goal. However,
several formulations require computationally demanding algorithms for posterior
inference. Motivated by this issue, we study a class of predictor-dependent
infinite mixture models, which relies on a simple representation of the
stick-breaking prior via sequential logistic regressions. This formulation
maintains the same desirable properties of popular predictor-dependent
stick-breaking priors, and leverages a recent P\'olya-gamma data augmentation
to facilitate the implementation of several computational methods for posterior
inference. These routines include Markov chain Monte Carlo via Gibbs sampling,
expectation-maximization algorithms, and mean-field variational Bayes for
scalable inference, thereby stimulating a wider implementation of Bayesian
density regression by practitioners. The algorithms associated with these
methods are presented in detail and tested in a toxicology study.
| stat |
A Multi-Way Correlation Coefficient | Pearson's correlation is an important summary measure of the amount of
dependence between two variables. It is natural to want to generalise the
concept of correlation as a single number that measures the inter-relatedness
of three or more variables e.g. how `correlated' are a collection of variables
in which non are specifically to be treated as an `outcome'? In this short
article, we introduce such a measure, and show that it reduces to the modulus
of Pearson's $r$ in the two dimensional case.
| stat |
Non-parametric estimation of Expectation and Variance of event count and
of incidence rate in a recurrent process -- where intensity of
event-occurrence changes with the occurrence of each higher order event | In this paper, a novel non-parametric method for estimation of expectation
and maximum value of the variance function is proposed for recurrent events
where intensity of event occurrence changes with the occurrence of each higher
order event. These kinds of recurrent events are often observed in clinical
trials for cardio-vascular events and also in many social experiments involving
drug addiction, armed robberies, etc. Simulated data is used to demonstrate the
novel approach for estimating the mean and variance of such recurrent events
and the results are compared with the result of Nelson Aalen estimator.
| stat |
Probabilistic Optimal Transport based on Collective Graphical Models | Optimal Transport (OT) is being widely used in various fields such as machine
learning and computer vision, as it is a powerful tool for measuring the
similarity between probability distributions and histograms. In previous
studies, OT has been defined as the minimum cost to transport probability mass
from one probability distribution to another. In this study, we propose a new
framework in which OT is considered as a maximum a posteriori (MAP) solution of
a probabilistic generative model. With the proposed framework, we show that OT
with entropic regularization is equivalent to maximizing a posterior
probability of a probabilistic model called Collective Graphical Model (CGM),
which describes aggregated statistics of multiple samples generated from a
graphical model. Interpreting OT as a MAP solution of a CGM has the following
two advantages: (i) We can calculate the discrepancy between noisy histograms
by modeling noise distributions. Since various distributions can be used for
noise modeling, it is possible to select the noise distribution flexibly to
suit the situation. (ii) We can construct a new method for interpolation
between histograms, which is an important application of OT. The proposed
method allows for intuitive modeling based on the probabilistic
interpretations, and a simple and efficient estimation algorithm is available.
Experiments using synthetic and real-world spatio-temporal population datasets
show the effectiveness of the proposed interpolation method.
| stat |
Differentially Private Precision Matrix Estimation | In this paper, we study the problem of precision matrix estimation when the
dataset contains sensitive information. In the differential privacy framework,
we develop a differentially private ridge estimator by perturbing the sample
covariance matrix. Then we develop a differentially private graphical lasso
estimator by using the alternating direction method of multipliers (ADMM)
algorithm. The theoretical results and empirical results that show the utility
of the proposed methods are also provided.
| stat |
A First Course in Data Science | Data science is a discipline that provides principles, methodology and
guidelines for the analysis of data for tools, values, or insights. Driven by a
huge workforce demand, many academic institutions have started to offer degrees
in data science, with many at the graduate, and a few at the undergraduate
level. Curricula may differ at different institutions, because of varying
levels of faculty expertise, and different disciplines (such as Math, computer
science, and business etc) in developing the curriculum. The University of
Massachusetts Dartmouth started offering degree programs in data science from
Fall 2015, at both the undergraduate and the graduate level. Quite a few
articles have been published that deal with graduate data science courses, much
less so dealing with undergraduate ones. Our discussion will focus on
undergraduate course structure and function, and specifically, a first course
in data science. Our design of this course centers around a concept called the
data science life cycle. That is, we view tasks or steps in the practice of
data science as forming a process, consisting of states that indicate how it
comes into life, how different tasks in data science depend on or interact with
others until the birth of a data product or the reach of a conclusion.
Naturally, different pieces of the data science life cycle then form individual
parts of the course. Details of each piece are filled up by concepts,
techniques, or skills that are popular in industry. Consequently, the design of
our course is both "principled" and practical. A significant feature of our
course philosophy is that, in line with activity theory, the course is based on
the use of tools to transform real data in order to answer strongly motivated
questions related to the data.
| stat |
A Symmetric Prior for Multinomial Probit Models | Fitted probabilities from widely used Bayesian multinomial probit models can
depend strongly on the choice of a base category, which is used to uniquely
identify the parameters of the model. This paper proposes a novel
identification strategy, and associated prior distribution for the model
parameters, that renders the prior symmetric with respect to relabeling the
outcome categories. The new prior permits an efficient Gibbs algorithm that
samples rank-deficient covariance matrices without resorting to
Metropolis-Hastings updates.
| stat |
System identification using Bayesian neural networks with nonparametric
noise models | System identification is of special interest in science and engineering. This
article is concerned with a system identification problem arising in stochastic
dynamic systems, where the aim is to estimating the parameters of a system
along with its unknown noise processes. In particular, we propose a Bayesian
nonparametric approach for system identification in discrete time nonlinear
random dynamical systems assuming only the order of the Markov process is
known. The proposed method replaces the assumption of Gaussian distributed
error components with a highly flexible family of probability density functions
based on Bayesian nonparametric priors. Additionally, the functional form of
the system is estimated by leveraging Bayesian neural networks which also leads
to flexible uncertainty quantification. Asymptotically on the number of hidden
neurons, the proposed model converges to full nonparametric Bayesian regression
model. A Gibbs sampler for posterior inference is proposed and its
effectiveness is illustrated in simulated and real time series.
| stat |
Ensemble Slice Sampling | Slice Sampling has emerged as a powerful Markov Chain Monte Carlo algorithm
that adapts to the characteristics of the target distribution with minimal
hand-tuning. However, Slice Sampling's performance is highly sensitive to the
user-specified initial length scale hyperparameter. Moreover, Slice Sampling
generally struggles with poorly scaled or strongly correlated distributions.
This paper introduces Ensemble Slice Sampling, a new class of algorithms that
bypasses such difficulties by adaptively tuning the length scale. Furthermore,
Ensemble Slice Sampling's performance is immune to linear correlations by
exploiting an ensemble of parallel walkers. These algorithms are trivial to
construct, require no hand-tuning, and can easily be implemented in parallel
computing environments. Empirical tests show that Ensemble Slice Sampling can
improve efficiency by more than an order of magnitude compared to conventional
MCMC methods on highly correlated target distributions such as the
Autoregressive Process of Order 1 and the Correlated Funnel distribution.
| stat |
A robust approach for ROC curves with covariates | The Receiver Operating Characteristic (ROC) curve is a useful tool that
measures the discriminating power of a continuous variable or the accuracy of a
pharmaceutical or medical test to distinguish between two conditions or
classes. In certain situations, the practitioner may be able to measure some
covariates related to the diagnostic variable which can increase the
discriminating power of the ROC curve. To protect against the existence of
atypical data among the observations, a procedure to obtain robust estimators
for the ROC curve in presence of covariates is introduced. The considered
proposal focusses on a semiparametric approach which fits a location-scale
regression model to the diagnostic variable and considers empirical estimators
of the regression residuals distributions. Robust parametric estimators are
combined with adaptive weighted empirical distribution estimators to
down-weight the influence of outliers. The uniform consistency of the proposal
is derived under mild assumptions. A Monte Carlo study is carried out to
compare the performance of the robust proposed estimators with the classical
ones both, in clean and contaminated samples. A real data set is also analysed.
| stat |
Discovering Clinically Meaningful Shape Features for the Analysis of
Tumor Pathology Images | With the advanced imaging technology, digital pathology imaging of tumor
tissue slides is becoming a routine clinical procedure for cancer diagnosis.
This process produces massive imaging data that capture histological details in
high resolution. Recent developments in deep-learning methods have enabled us
to automatically detect and characterize the tumor regions in pathology images
at large scale. From each identified tumor region, we extracted 30 well-defined
descriptors that quantify its shape, geometry, and topology. We demonstrated
how those descriptor features were associated with patient survival outcome in
lung adenocarcinoma patients from the National Lung Screening Trial (n=143).
Besides, a descriptor-based prognostic model was developed and validated in an
independent patient cohort from The Cancer Genome Atlas Program program
(n=318). This study proposes new insights into the relationship between tumor
shape, geometrical, and topological features and patient prognosis. We provide
software in the form of R code on GitHub:
https://github.com/estfernandez/Slide_Image_Segmentation_and_Extraction.
| stat |
Approximation results regarding the multiple-output mixture of linear
experts model | Mixture of experts (MoE) models are a class of artificial neural networks
that can be used for functional approximation and probabilistic modeling. An
important class of MoE models is the class of mixture of linear experts (MoLE)
models, where the expert functions map to real topological output spaces. There
are a number of powerful approximation results regarding MoLE models, when the
output space is univariate. These results guarantee the ability of MoLE mean
functions to approximate arbitrary continuous functions, and MoLE models
themselves to approximate arbitrary conditional probability density functions.
We utilize and extend upon the univariate approximation results in order to
prove a pair of useful results for situations where the output spaces are
multivariate.
| stat |
Time-to-event regression using partially monotonic neural networks | We propose a novel method, termed SuMo-net, that uses partially monotonic
neural networks to learn a time-to-event distribution from a sample of
covariates and right-censored times. SuMo-net models the survival function and
the density jointly, and optimizes the likelihood for right-censored data
instead of the often used partial likelihood. The method does not make
assumptions about the true survival distribution and avoids computationally
expensive integration of the hazard function. We evaluate the performance of
the method on a range of datasets and find competitive performance across
different metrics and improved computational time of making new predictions.
| stat |
next-gen-scraPy: Extracting NFL Tracking Data from Images to Evaluate
Quarterbacks and Pass Defenses | The NFL collects detailed tracking data capturing the location of all players
and the ball during each play. Although the raw form of this data is not
publicly available, the NFL releases a set of aggregated statistics via their
Next Gen Stats (NGS) platform. They also provide charts showing the locations
of pass attempts and outcomes for individual quarterbacks. Our work aims to
partially close the gap between what data is available privately (to NFL teams)
and publicly, and our contribution is twofold. First, we introduce an image
processing tool designed specifically for extracting the raw data from the NGS
pass charts. We extract the pass outcome, coordinates, and other metadata.
Second, we analyze the resulting dataset, examining the spatial tendencies and
performances of individual quarterbacks and defenses. We use a generalized
additive model for completion percentages by field location. We introduce a
Naive Bayes approach for estimating the 2-D completion percentage surfaces of
individual teams and quarterbacks, and we provide a one-number summary,
completion percentage above expectation (CPAE), for evaluating quarterbacks and
team defenses. We find that our pass location data closely matches the NFL's
tracking data, and that our CPAE metric closely matches the NFL's proprietary
CPAE metric.
| stat |
Bayesian Nonparametric Spectral Estimation | Spectral estimation (SE) aims to identify how the energy of a signal (e.g., a
time series) is distributed across different frequencies. This can become
particularly challenging when only partial and noisy observations of the signal
are available, where current methods fail to handle uncertainty appropriately.
In this context, we propose a joint probabilistic model for signals,
observations and spectra, where SE is addressed as an exact inference problem.
Assuming a Gaussian process prior over the signal, we apply Bayes' rule to find
the analytic posterior distribution of the spectrum given a set of
observations. Besides its expressiveness and natural account of spectral
uncertainty, the proposed model also provides a functional-form representation
of the power spectral density, which can be optimised efficiently. Comparison
with previous approaches, in particular against Lomb-Scargle, is addressed
theoretically and also experimentally in three different scenarios. Code and
demo available at https://github.com/GAMES-UChile/BayesianSpectralEstimation.
| stat |
Estimation and Inference on Heterogeneous Treatment Effects in
High-Dimensional Dynamic Panels | This paper provides estimation and inference methods for a large number of
heterogeneous treatment effects in the presence of an even larger number of
controls and unobserved unit heterogeneity. In our main example, the vector of
heterogeneous treatments is generated by interacting the base treatment
variable with a subset of controls. We first estimate the unit-specific
expectation functions of the outcome and each treatment interaction conditional
on controls and take the residuals. Second, we report the Lasso (L1-regularized
least squares) estimate of the heterogeneous treatment effect parameter,
regressing the outcome residual on the vector of treatment ones. We debias the
Lasso estimator to conduct simultaneous inference on the target parameter by
Gaussian bootstrap. We account for the unobserved unit heterogeneity by
projecting it onto the time-invariant covariates, following the correlated
random effects approach of Mundlak (1978) and Chamberlain (1982). We
demonstrate our method by estimating price elasticities of groceries based on
scanner data.
| stat |
On The Problem of Relevance in Statistical Inference | This paper is dedicated to the "50 Years of the Relevance Problem" - a
long-neglected topic that begs attention from practical statisticians who are
concerned with the problem of drawing inference from large-scale heterogeneous
data.
| stat |
Bayesian Gamma-Negative Binomial Modeling of Single-Cell RNA Sequencing
Data | Background: Single-cell RNA sequencing (scRNA-seq) is a powerful profiling
technique at the single-cell resolution. Appropriate analysis of scRNA-seq data
can characterize molecular heterogeneity and shed light into the underlying
cellular process to better understand development and disease mechanisms. The
unique analytic challenge is to appropriately model highly over-dispersed
scRNA-seq count data with prevalent dropouts (zero counts), making
zero-inflated dimensionality reduction techniques popular for scRNA-seq data
analyses. Employing zero-inflated distributions, however, may place extra
emphasis on zero counts, leading to potential bias when identifying the latent
structure of the data. Results: In this paper, we propose a fully generative
hierarchical gamma-negative binomial (hGNB) model of scRNA-seq data, obviating
the need for explicitly modeling zero inflation. At the same time, hGNB can
naturally account for covariate effects at both the gene and cell levels to
identify complex latent representations of scRNA-seq data, without the need for
commonly adopted pre-processing steps such as normalization. Efficient Bayesian
model inference is derived by exploiting conditional conjugacy via novel data
augmentation techniques. Conclusion: Experimental results on both simulated
data and several real-world scRNA-seq datasets suggest that hGNB is a powerful
tool for cell cluster discovery as well as cell lineage inference.
| stat |
Quantile Regression Modelling via Location and Scale Mixtures of Normal
Distributions | We show that the estimating equations for quantile regression can be solved
using a simple EM algorithm in which the M-step is computed via weighted least
squares, with weights computed at the E-step as the expectation of independent
generalized inverse-Gaussian variables. We compute the variance-covariance
matrix for the quantile regression coefficients using a kernel density
estimator that results in more stable standard errors than those produced by
existing software. A natural modification of the EM algorithm that involves
fitting a linear mixed model at the M-step extends the methodology to mixed
effects quantile regression models. In this case, the fitting method can be
justified as a generalized alternating minimization algorithm. Obtaining
quantile regression estimates via the weighted least squares method enables
model diagnostic techniques similar to the ones used in the linear regression
setting. The computational approach is compared with existing software using
simulated data, and the methodology is illustrated with several case studies.
| stat |
Bayesian joint modeling of chemical structure and dose response curves | Today there are approximately 85,000 chemicals regulated under the Toxic
Substances Control Act, with around 2,000 new chemicals introduced each year.
It is impossible to screen all of these chemicals for potential toxic effects
either via full organism in vivo studies or in vitro high-throughput screening
(HTS) programs. Toxicologists face the challenge of choosing which chemicals to
screen, and predicting the toxicity of as-yet-unscreened chemicals. Our goal is
to describe how variation in chemical structure relates to variation in
toxicological response to enable in silico toxicity characterization designed
to meet both of these challenges. With our Bayesian partially Supervised Sparse
and Smooth Factor Analysis ($\text{BS}^3\text{FA}$) model, we learn a distance
between chemicals targeted to toxicity, rather than one based on molecular
structure alone. Our model also enables the prediction of chemical
dose-response profiles based on chemical structure (that is, without in vivo or
in vitro testing) by taking advantage of a large database of chemicals that
have already been tested for toxicity in HTS programs. We show superior
simulation performance in distance learning and modest to large gains in
predictive ability compared to existing methods. Results from the
high-throughput screening data application elucidate the relationship between
chemical structure and a toxicity-relevant high-throughput assay. An R package
for $\text{BS}^3\text{FA}$ is available online at
https://github.com/kelrenmor/bs3fa.
| stat |
Multivariate Log-Contrast Regression with Sub-Compositional Predictors:
Testing the Association Between Preterm Infants' Gut Microbiome and
Neurobehavioral Outcomes | The so-called gut-brain axis has stimulated extensive research on
microbiomes. One focus is to assess the association between certain clinical
outcomes and the relative abundances of gut microbes, which can be presented as
sub-compositional data in conformity with the taxonomic hierarchy of bacteria.
Motivated by a study for identifying the microbes in the gut microbiome of
preterm infants that impact their later neurobehavioral outcomes, we formulate
a constrained integrative multi-view regression, where the neurobehavioral
scores form multivariate response, the sub-compositional microbiome data form
multi-view feature matrices, and a set of linear constraints on their
corresponding sub-coefficient matrices ensures the conformity to the simplex
geometry. To enable joint selection and inference of sub-compositions/views, we
assume all the sub-coefficient matrices are possibly of low-rank, i.e., the
outcomes are associated with the microbiome through different sets of latent
sub-compositional factors from different taxa. We propose a scaled composite
nuclear norm penalization approach for model estimation and develop a
hypothesis testing procedure through de-biasing to assess the significance of
different views. Simulation studies confirm the effectiveness of the proposed
procedure. In the preterm infant study, the identified microbes are mostly
consistent with existing studies and biological understandings. Our approach
supports that stressful early life experiences imprint gut microbiome through
the regulation of the gut-brain axis.
| stat |
Support recovery and sup-norm convergence rates for sparse pivotal
estimation | In high dimensional sparse regression, pivotal estimators are estimators for
which the optimal regularization parameter is independent of the noise level.
The canonical pivotal estimator is the square-root Lasso, formulated along with
its derivatives as a "non-smooth + non-smooth" optimization problem. Modern
techniques to solve these include smoothing the datafitting term, to benefit
from fast efficient proximal algorithms. In this work we show minimax sup-norm
convergence rates for non smoothed and smoothed, single task and multitask
square-root Lasso-type estimators. Thanks to our theoretical analysis, we
provide some guidelines on how to set the smoothing hyperparameter, and
illustrate on synthetic data the interest of such guidelines.
| stat |
The Elliptical Potential Lemma Revisited | This note proposes a new proof and new perspectives on the so-called
Elliptical Potential Lemma. This result is important in online learning,
especially for linear stochastic bandits. The original proof of the result,
however short and elegant, does not give much flexibility on the type of
potentials considered and we believe that this new interpretation can be of
interest for future research in this field.
| stat |
Bi-factor and second-order copula models for item response data | Bi-factor and second-order models based on copulas are proposed for item
response data, where the items can be split into non-overlapping groups such
that there is a homogeneous dependence within each group. Our general models
include the Gaussian bi-factor and second-order models as special cases and can
lead to more probability in the joint upper or lower tail compared with the
Gaussian bi-factor and second-order models. Details on maximum likelihood
estimation of parameters for the bi-factor and second-order copula models are
given, as well as model selection and goodness-of-fit techniques. Our general
methodology is demonstrated with an extensive simulation study and illustrated
for the Toronto Alexithymia Scale. Our studies suggest that there can be a
substantial improvement over the Gaussian bi-factor and second-order models
both conceptually, as the items can have interpretations of latent
maxima/minima or mixtures of means in comparison with latent means, and in fit
to data.
| stat |
Incorporation of frailties into a non-proportional hazard regression
model and its diagnostics for reliability modeling of downhole safety valves | In this paper, our proposal consists of incorporating frailty into a
statistical methodology for modeling time-to-event data, based on
non-proportional hazards regression model. Specifically, we use the generalized
time-dependent logistic (GTDL) model with a frailty term introduced in the
hazard function to control for unobservable heterogeneity among the sampling
units. We also add a regression in the parameter that measures the effect of
time, since it can directly reflect the influence of covariates on the effect
of time-to-failure. The practical relevance of the proposed model is
illustrated in a real problem based on a data set for downhole safety valves
(DHSVs) used in offshore oil and gas production wells. The reliability
estimation of DHSVs can be used, among others, to predict the blowout
occurrence, assess the workover demand and aid decision-making actions.
| stat |
gfoRmula: An R package for estimating effects of general time-varying
treatment interventions via the parametric g-formula | Researchers are often interested in using longitudinal data to estimate the
causal effects of hypothetical time-varying treatment interventions on the mean
or risk of a future outcome. Standard regression/conditioning methods for
confounding control generally fail to recover causal effects when time-varying
confounders are themselves affected by past treatment. In such settings,
estimators derived from Robins's g-formula may recover time-varying treatment
effects provided sufficient covariates are measured to control confounding by
unmeasured risk factors. The package gfoRmula implements in R one such
estimator: the parametric g-formula. This estimator easily adapts to binary or
continuous time-varying treatments as well as contrasts defined by static or
dynamic, deterministic or random treatment interventions, as well as
interventions that depend on the natural value of treatment. The package
accommodates survival outcomes as well as binary or continuous end of follow-up
outcomes. For survival outcomes, the package has different options for handling
competing events. This paper describes the gfoRmula package, along with
motivating background, features, and examples.
| stat |
On Benjamini-Hochberg procedure applied to mid p-values | Multiple testing with discrete p-values routinely arises in various
scientific endeavors. However, procedures, including the false discovery rate
(FDR) controlling Benjamini-Hochberg (BH) procedure, often used in such
settings, being developed originally for p-values with continuous
distributions, are too conservative, and so may not be as powerful as one would
hope for. Therefore, improving the BH procedure by suitably adapting it to
discrete p-values without losing its FDR control is currently an important path
of research. This paper studies the FDR control of the BH procedure when it is
applied to mid p-values and derive conditions under which it is conservative.
Our simulation study reveals that the BH procedure applied to mid p-values may
be conservative under much more general settings than characterized in this
work, and that an adaptive version of the BH procedure applied to mid p-values
is as powerful as an existing adaptive procedure based on randomized p-values.
| stat |
When OT meets MoM: Robust estimation of Wasserstein Distance | Issued from Optimal Transport, the Wasserstein distance has gained importance
in Machine Learning due to its appealing geometrical properties and the
increasing availability of efficient approximations. In this work, we consider
the problem of estimating the Wasserstein distance between two probability
distributions when observations are polluted by outliers. To that end, we
investigate how to leverage Medians of Means (MoM) estimators to robustify the
estimation of Wasserstein distance. Exploiting the dual Kantorovitch
formulation of Wasserstein distance, we introduce and discuss novel MoM-based
robust estimators whose consistency is studied under a data contamination model
and for which convergence rates are provided. These MoM estimators enable to
make Wasserstein Generative Adversarial Network (WGAN) robust to outliers, as
witnessed by an empirical study on two benchmarks CIFAR10 and Fashion MNIST.
Eventually, we discuss how to combine MoM with the entropy-regularized
approximation of the Wasserstein distance and propose a simple MoM-based
re-weighting scheme that could be used in conjunction with the Sinkhorn
algorithm.
| stat |
ROCnReg: An R Package for Receiver Operating Characteristic Curve
Inference with and without Covariate Information | The receiver operating characteristic (ROC) curve is the most popular tool
used to evaluate the discriminatory capability of diagnostic tests/biomarkers
measured on a continuous scale when distinguishing between two alternative
disease states (e.g, diseased and nondiseased). In some circumstances, the
test's performance and its discriminatory ability may vary according to
subject-specific characteristics or different test settings. In such cases,
information-specific accuracy measures, such as the covariate-specific and the
covariate-adjusted ROC curve are needed, as ignoring covariate information may
lead to biased or erroneous results. This paper introduces the R package
ROCnReg that allows estimating the pooled (unadjusted) ROC curve, the
covariate-specific ROC curve, and the covariate-adjusted ROC curve by different
methods, both from (semi) parametric and nonparametric perspectives and within
Bayesian and frequentist paradigms. From the estimated ROC curve (pooled,
covariate-specific or covariate-adjusted), several summary measures of
accuracy, such as the (partial) area under the ROC curve and the Youden index,
can be obtained. The package also provides functions to obtain ROC-based
optimal threshold values using several criteria, namely, the Youden Index
criterion and the criterion that sets a target value for the false positive
fraction. For the Bayesian methods, we provide tools for assessing model fit
via posterior predictive checks, while model choice can be carried out via
several information criteria. Numerical and graphical outputs are provided for
all methods. The package is illustrated through the analyses of data from an
endocrine study where the aim is to assess the capability of the body mass
index to detect the presence or absence of cardiovascular disease risk factors.
The package is available from CRAN at
https://CRAN.R-project.org/package=ROCnReg.
| stat |
Estimating heterogeneous wildfire effects using synthetic controls and
satellite remote sensing | Wildfires have become one of the biggest natural hazards for environments
worldwide. The effects of wildfires are heterogeneous, meaning that the
magnitude of their effects depends on many factors such as geographical region,
climate and land cover/vegetation type. Yet, which areas are more affected by
these events remains unclear. Here we present a novel application of the
Generalised Synthetic Control (GSC) method that enables quantification and
prediction of vegetation changes due to wildfires through a time-series
analysis of in situ and satellite remote sensing data. We apply this method to
medium to large wildfires ($>$ 1000 acres) in California throughout a time-span
of two decades (1996--2016). The method's ability for estimating counterfactual
vegetation characteristics for burned regions is explored in order to quantify
abrupt system changes. We find that the GSC method is better at predicting
vegetation changes than the more traditional approach of using nearby regions
to assess wildfire impacts. We evaluate the GSC method by comparing its
predictions of spectral vegetation indices to observations during pre-wildfire
periods and find improvements in correlation coefficient from $R^2 = 0.66$ to
$R^2 = 0.93$ in Normalised Difference Vegetation Index (NDVI), from $R^2 =
0.48$ to $R^2 = 0.81$ for Normalised Burn Ratio (NBR), and from $R^2 = 0.49$ to
$R^2 = 0.85$ for Normalised Difference Moisture Index (NDMI). Results show
greater changes in NDVI, NBR, and NDMI post-fire on regions classified as
having a lower Burning Index. The GSC method also reveals that wildfire effects
on vegetation can last for more than a decade post-wildfire, and in some cases
never return to their previous vegetation cycles within our study period.
Lastly, we discuss the usefulness of using GSC in remote sensing analyses.
| stat |
Novel Uncertainty Framework for Deep Learning Ensembles | Deep neural networks have become the default choice for many of the machine
learning tasks such as classification and regression. Dropout, a method
commonly used to improve the convergence of deep neural networks, generates an
ensemble of thinned networks with extensive weight sharing. Recent studies that
dropout can be viewed as an approximate variational inference in Gaussian
processes, and used as a practical tool to obtain uncertainty estimates of the
network. We propose a novel statistical mechanics based framework to dropout
and use this framework to propose a new generic algorithm that focuses on
estimates of the variance of the loss as measured by the ensemble of thinned
networks. Our approach can be applied to a wide range of deep neural network
architectures and machine learning tasks. In classification, this algorithm
allows the generation of a don't-know answer to be generated, which can
increase the reliability of the classifier. Empirically we demonstrate
state-of-the-art AUC results on publicly available benchmarks.
| stat |
Bayesian Kernelised Test of (In)dependence with Mixed-type Variables | A fundamental task in AI is to assess (in)dependence between mixed-type
variables (text, image, sound). We propose a Bayesian kernelised correlation
test of (in)dependence using a Dirichlet process model. The new measure of
(in)dependence allows us to answer some fundamental questions: Based on data,
are (mixed-type) variables independent? How likely is dependence/independence
to hold? How high is the probability that two mixed-type variables are more
than just weakly dependent? We theoretically show the properties of the
approach, as well as algorithms for fast computation with it. We empirically
demonstrate the effectiveness of the proposed method by analysing its
performance and by comparing it with other frequentist and Bayesian approaches
on a range of datasets and tasks with mixed-type variables.
| stat |
Graphical outputs and Spatial Cross-validation for the R-INLA package
using INLAutils | Statistical analyses proceed by an iterative process of model fitting and
checking. The R-INLA package facilitates this iteration by fitting many
Bayesian models much faster than alternative MCMC approaches. As the
interpretation of results and model objects from Bayesian analyses can be
complex, the R package INLAutils provides users with easily accessible, clear
and customisable graphical summaries of model outputs from R- INLA.
Furthermore, it offers a function for performing and visualizing the results of
a spatial leave-one-out cross-validation (SLOOCV) approach that can be applied
to compare the predictive performance of multiple spatial models. In this
paper, we describe and illustrate the use of (1) graphical summary plotting
functions and (2) the SLOOCV approach. We conclude the paper by identifying the
limits of our approach and discuss future potential improvements.
| stat |
Generalizing Randomized Trial Findings to a Target Population using
Complex Survey Population Data | Randomized trials are considered the gold standard for estimating causal
effects. Trial findings are often used to inform policy and programming
efforts, yet their results may not generalize well to a relevant target
population due to potential differences in effect moderators between the trial
and population. Statistical methods have been developed to improve
generalizability by combining trials and population data, and weighting the
trial to resemble the population on baseline covariates.Large-scale surveys in
fields such as health and education with complex survey designs are a logical
source for population data; however, there is currently no best practice for
incorporating survey weights when generalizing trial findings to a complex
survey. We propose and investigate ways to incorporate survey weights in this
context. We examine the performance of our proposed estimator in simulations by
comparing its performance to estimators that ignore the complex survey
design.We then apply the methods to generalize findings from two trials - a
lifestyle intervention for blood pressure reduction and a web-based
intervention to treat substance use disorders - to their respective target
populations using population data from complex surveys. The work highlights the
importance in properly accounting for the complex survey design when
generalizing trial findings to a population represented by a complex survey
sample.
| stat |
Perfect Gibbs Sampling of Order Constrained Non-IID Ordered Random
Variates with Application to Bayesian Principal Components Analysis | Order statistics arising from $m$ independent but not identically distributed
random variables are typically constructed by arranging some $X_{1}, X_{2},
\ldots, X_{m}$, with $X_{i}$ having distribution function $F_{i}(x)$, in
increasing order denoted as $X_{(1)} \leq X_{(2)} \leq \ldots \leq X_{(m)}$. In
this case, $X_{(i)}$ is not necessarily associated with $F_{i}(x)$. Assuming
one can simulate values from each distribution, one can generate such
``non-iid" order statistics by simulating $X_{i}$ from $F_{i}$, for
$i=1,2,\ldots, m$, and simply putting them in order. In this paper, we consider
the problem of simulating ordered values $X_{(1)}, X_{(2)}, \ldots, X_{(m)}$
such that the marginal distribution of $X_{(i)}$ is $F_{i}(x)$. This problem
arises in Bayesian principal components analysis (BPCA) where the $X_{i}$ are
ordered eigenvalues that are a posteriori independent but not identically
distributed. In this paper, we propose a novel {\emph{coupling-from-the-past}}
algorithm to ``perfectly" (up to computable order of accuracy) simulate such
"order-constrained non-iid" order statistics. We demonstrate the effectiveness
of our approach for several examples, including the BPCA problem.
| stat |
Advanced Algorithms for Penalized Quantile and Composite Quantile
Regression | In this paper, we discuss a family of robust, high-dimensional regression
models for quantile and composite quantile regression, both with and without an
adaptive lasso penalty for variable selection. We reformulate these quantile
regression problems and obtain estimators by applying the alternating direction
method of multipliers (ADMM), majorize-minimization (MM), and coordinate
descent (CD) algorithms. Our new approaches address the lack of publicly
available methods for (composite) quantile regression, especially for
high-dimensional data, both with and without regularization. Through simulation
studies, we demonstrate the need for different algorithms applicable to a
variety of data settings, which we implement in the cqrReg package for R. For
comparison, we also introduce the widely used interior point (IP) formulation
and test our methods against the IP algorithms in the existing quantreg
package. Our simulation studies show that each of our methods, particularly MM
and CD, excel in different settings such as with large or high-dimensional data
sets, respectively, and outperform the methods currently implemented in
quantreg. The ADMM approach offers specific promise for future developments in
its amenability to parallelization and scalability.
| stat |
Learning a latent pattern of heterogeneity in the innovation rates of a
time series of counts | We develop a Bayesian hierarchical semiparametric model for phenomena related
to time series of counts. The main feature of the model is its capability to
learn a latent pattern of heterogeneity in the distribution of the process
innovation rates, which are softly clustered through time with the help of a
Dirichlet process placed at the top of the model hierarchy. The probabilistic
forecasting capabilities of the model are put to test in the analysis of crime
data in Pittsburgh, with favorable results.
| stat |
Order selection with confidence for finite mixture models | The determination of the number of mixture components (the order) of a finite
mixture model has been an enduring problem in statistical inference. We prove
that the closed testing principle leads to a sequential testing procedure (STP)
that allows for confidence statements to be made regarding the order of a
finite mixture model. We construct finite sample tests, via data splitting and
data swapping, for use in the STP, and we prove that such tests are consistent
against fixed alternatives. Simulation studies are conducted to demonstrate the
performance of the finite sample tests-based STP, yielding practical
recommendations, and extensions to the STP are considered. In particular, we
demonstrate that a modification of the STP yields a method that consistently
selects the order of a finite mixture model, in the asymptotic sense. Our STP
not only applicable for order selection of finite mixture models, but is also
useful for making confidence statements regarding any sequence of nested
models.
| stat |
Intrinsic Riemannian Functional Data Analysis for Sparse Longitudinal
Observations | A novel framework is developed to intrinsically analyze sparsely observed
Riemannian functional data. It features four innovative components: a
frame-independent covariance function, a smooth vector bundle termed covariance
vector bundle, a parallel transport and a smooth bundle metric on the
covariance vector bundle. The introduced intrinsic covariance function links
estimation of covariance structure to smoothing problems that involve raw
covariance observations derived from sparsely observed Riemannian functional
data, while the covariance vector bundle provides a rigorous mathematical
foundation for formulating the smoothing problems. The parallel transport and
the bundle metric together make it possible to measure fidelity of fit to the
covariance function. They also plays a critical role in quantifying the quality
of estimators for the covariance function. As an illustration, based on the
proposed framework, we develop a local linear smoothing estimator for the
covariance function, analyze its theoretical properties, and provide numerical
demonstration via simulated and real datasets. The intrinsic feature of the
framework makes it applicable to not only Euclidean submanifolds but also
manifolds without a canonical ambient space.
| stat |
GPMatch: A Bayesian Doubly Robust Approach to Causal Inference with
Gaussian Process Covariance Function As a Matching Tool | Gaussian process (GP) covariance function is proposed as a matching tool in
GPMatch within a full Bayesian framework under relatively weaker causal
assumptions. The matching is accomplished by utilizing GP prior covariance
function to define matching distance. We show that GPMatch provides a doubly
robust estimate of the averaged treatment effect (ATE) much like the
G-estimation, the ATE is correctly estimated when either conditions are
satisfied: 1) the GP mean function correctly specifies potential outcome
\(Y^{(0)}\); or 2) the GP covariance function correctly specifies matching
structure. Simulation studies were carried out without assuming any known
matching structure nor functional form of the outcomes. The results demonstrate
that GPMatch enjoys well calibrated frequentist properties, and outperforms
many widely used methods including Bayesian Additive Regression Trees. The case
study compares effectiveness of early aggressive use of biological medication
in treating children with newly diagnosed Juvenile Idiopathic Arthritis, using
data extracted from electronic medical records.
| stat |
Horseshoe shrinkage methods for Bayesian fusion estimation | We consider the problem of estimation and structure learning of high
dimensional signals via a normal sequence model, where the underlying parameter
vector is piecewise constant, or has a block structure. We develop a Bayesian
fusion estimation method by using the Horseshoe prior to induce a strong
shrinkage effect on successive differences in the mean parameters,
simultaneously imposing sufficient prior concentration for non-zero values of
the same. The proposed method thus facilitates consistent estimation and
structure recovery of the signal pieces. We provide theoretical justifications
of our approach by deriving posterior convergence rates and establishing
selection consistency under suitable assumptions. We also extend our proposed
method to signal de-noising over arbitrary graphs and develop efficient
computational methods along with providing theoretical guarantees. We
demonstrate the superior performance of the Horseshoe based Bayesian fusion
estimation method through extensive simulations and two real-life examples on
signal de-noising in biological and geophysical applications. We also
demonstrate the estimation performance of our method on a real-world large
network for the graph signal de-noising problem.
| stat |
Explaining the difference between men's and women's football | Women's football is gaining supporters and practitioners worldwide, raising
questions about what the differences are with men's football. While the two
sports are often compared based on the players' physical attributes, we analyze
the spatio-temporal events during matches in the last World Cups to compare
male and female teams based on their technical performance. We train an
artificial intelligence model to recognize if a team is male or female based on
variables that describe a match's playing intensity, accuracy, and performance
quality. Our model accurately distinguishes between men's and women's football,
revealing crucial technical differences, which we investigate through the
extraction of explanations from the classifier's decisions. The differences
between men's and women's football are rooted in play accuracy, the recovery
time of ball possession, and the players' performance quality. Our methodology
may help journalists and fans understand what makes women's football a distinct
sport and coaches design tactics tailored to female teams.
| stat |
Unbiased Markov chain Monte Carlo for intractable target distributions | Performing numerical integration when the integrand itself cannot be
evaluated point-wise is a challenging task that arises in statistical analysis,
notably in Bayesian inference for models with intractable likelihood functions.
Markov chain Monte Carlo (MCMC) algorithms have been proposed for this setting,
such as the pseudo-marginal method for latent variable models and the exchange
algorithm for a class of undirected graphical models. As with any MCMC
algorithm, the resulting estimators are justified asymptotically in the limit
of the number of iterations, but exhibit a bias for any fixed number of
iterations due to the Markov chains starting outside of stationarity. This
"burn-in" bias is known to complicate the use of parallel processors for MCMC
computations. We show how to use coupling techniques to generate unbiased
estimators in finite time, building on recent advances for generic MCMC
algorithms. We establish the theoretical validity of some of these procedures
by extending existing results to cover the case of polynomially ergodic Markov
chains. The efficiency of the proposed estimators is compared with that of
standard MCMC estimators, with theoretical arguments and numerical experiments
including state space models and Ising models.
| stat |
Data-generating models under which the random forest algorithm performs
badly | Examples are given of data-generating models under which some versions of the
random forest algorithm may fail to be consistent or be extremely slow to
converge to the optimal predictor. The evidence provided for these properties
is based on mostly intuitive arguments, similar to those used earlier with
simpler examples, and on numerical experiments. Although one can always choose
a model under which random forests perform very badly, it is shown that when
substantial improvement is possible simple methods based on statistics of
'variable use' and 'variable importance' may indicate a better predictor based
on a sort of mixture of random forests; thus, by acknowledging the difficulties
posed by some models one may improve the performance of random forests in some
applications.
| stat |
forgeNet: A graph deep neural network model using tree-based ensemble
classifiers for feature extraction | A unique challenge in predictive model building for omics data has been the
small number of samples $(n)$ versus the large amount of features $(p)$. This
"$n\ll p$" property brings difficulties for disease outcome classification
using deep learning techniques. Sparse learning by incorporating external gene
network information such as the graph-embedded deep feedforward network (GEDFN)
model has been a solution to this issue. However, such methods require an
existing feature graph, and potential mis-specification of the feature graph
can be harmful on classification and feature selection. To address this
limitation and develop a robust classification model without relying on
external knowledge, we propose a \underline{for}est
\underline{g}raph-\underline{e}mbedded deep feedforward \underline{net}work
(forgeNet) model, to integrate the GEDFN architecture with a forest feature
graph extractor, so that the feature graph can be learned in a supervised
manner and specifically constructed for a given prediction task. To validate
the method's capability, we experimented the forgeNet model with both synthetic
and real datasets. The resulting high classification accuracy suggests that the
method is a valuable addition to sparse deep learning models for omics data.
| stat |
Decision Making in Drug Development via Confidence Distributions | In clinical drug development a typical phase three power calculation for a
Go/No-Go decision is performed by replacing unknown population-level quantities
in the power function with what is observed from a literature review or what is
observed in phase two. Many authors and practitioners view this as an assumed
value of power and offer the Bayesian quantity probability of success or
assurance as an alternative. The claim is by averaging over a prior or
posterior distribution, probability of success transcends power by capturing
the uncertainty around the unknown true treatment effect and any other
population-level parameters. We use confidence distributions to frame both the
probability of success calculation and the typical power calculation as merely
producing two different point estimates of power. We demonstrate that Go/No-Go
decisions based on either point estimate of power do not adequately quantify
and control the risk involved, and instead we argue for Go/No-Go decisions that
utilize inference on power for better risk management and decision making. This
inference on power can be derived and displayed using confidence distributions.
| stat |
Fairness constraints can help exact inference in structured prediction | Many inference problems in structured prediction can be modeled as maximizing
a score function on a space of labels, where graphs are a natural
representation to decompose the total score into a sum of unary (nodes) and
pairwise (edges) scores. Given a generative model with an undirected connected
graph $G$ and true vector of binary labels, it has been previously shown that
when $G$ has good expansion properties, such as complete graphs or $d$-regular
expanders, one can exactly recover the true labels (with high probability and
in polynomial time) from a single noisy observation of each edge and node. We
analyze the previously studied generative model by Globerson et al. (2015)
under a notion of statistical parity. That is, given a fair binary node
labeling, we ask the question whether it is possible to recover the fair
assignment, with high probability and in polynomial time, from single edge and
node observations. We find that, in contrast to the known trade-offs between
fairness and model performance, the addition of the fairness constraint
improves the probability of exact recovery. We effectively explain this
phenomenon and empirically show how graphs with poor expansion properties, such
as grids, are now capable to achieve exact recovery with high probability.
Finally, as a byproduct of our analysis, we provide a tighter
minimum-eigenvalue bound than that of Weyl's inequality.
| stat |
Robust data-driven discovery of partial differential equations with
time-dependent coefficients | In this work, we propose a robust Bayesian sparse learning algorithm based on
Bayesian group Lasso with spike and slab priors for the discovery of partial
differential equations with variable coefficients. Using the samples draw from
the posterior distribution with a Gibbs sampler, we are able to estimate the
values of coefficients, together with their standard errors and confidence
intervals. Apart from constructing the error bars, uncertainty quantification
can also be employed for designing new criteria of model selection and
threshold setting. This enables our method more adjustable and robust in
learning equations with time-dependent coefficients. Three criteria are
introduced for model selection and threshold setting to identify the correct
terms: the root mean square, total error bar, and group error bar. Moreover,
three noise filters are integrated with the robust Bayesian sparse learning
algorithm for better results with larger noise. Numerical results demonstrate
that our method is more robust than sequential grouped threshold ridge
regression and group Lasso in noisy situations through three examples.
| stat |
Optimal designs for model averaging in non-nested models | In this paper we construct optimal designs for frequentist model averaging
estimation. We derive the asymptotic distribution of the model averaging
estimate with fixed weights in the case where the competing models are
non-nested and none of these models is correctly specified. A Bayesian optimal
design minimizes an expectation of the asymptotic mean squared error of the
model averaging estimate calculated with respect to a suitable prior
distribution. We demonstrate that Bayesian optimal designs can improve the
accuracy of model averaging substantially. Moreover, the derived designs also
improve the accuracy of estimation in a model selected by model selection and
model averaging estimates with random weights.
| stat |
A Kernel-Based Neural Network for High-dimensional Genetic Risk
Prediction Analysis | Risk prediction capitalizing on emerging human genome findings holds great
promise for new prediction and prevention strategies. While the large amounts
of genetic data generated from high-throughput technologies offer us a unique
opportunity to study a deep catalog of genetic variants for risk prediction,
the high-dimensionality of genetic data and complex relationships between
genetic variants and disease outcomes bring tremendous challenges to risk
prediction analysis. To address these rising challenges, we propose a
kernel-based neural network (KNN) method. KNN inherits features from both
linear mixed models (LMM) and classical neural networks and is designed for
high-dimensional risk prediction analysis. To deal with datasets with millions
of variants, KNN summarizes genetic data into kernel matrices and use the
kernel matrices as inputs. Based on the kernel matrices, KNN builds a
single-layer feedforward neural network, which makes it feasible to consider
complex relationships between genetic variants and disease outcomes. The
parameter estimation in KNN is based on MINQUE and we show, that under certain
conditions, the average prediction error of KNN can be smaller than that of
LMM. Simulation studies also confirm the results.
| stat |
Sparse estimation for case-control studies with multiple subtypes of
cases | The analysis of case-control studies with several subtypes of cases is
increasingly common, e.g. in cancer epidemiology. For matched designs, we show
that a natural strategy is based on a stratified conditional logistic
regression model. Then, to account for the potential homogeneity among the
subtypes of cases, we adapt the ideas of data shared lasso, which has been
recently proposed for the estimation of regression models in a stratified
setting. For unmatched designs, we compare two standard methods based on
L1-norm penalized multinomial logistic regression. We describe formal
connections between these two approaches, from which practical guidance can be
derived. We show that one of these approaches, which is based on a symmetric
formulation of the multinomial logistic regression model, actually reduces to a
data shared lasso version of the other. Consequently, the relative performance
of the two approaches critically depends on the level of homogeneity that
exists among the subtypes of cases: more precisely, when homogeneity is
moderate to high, the non-symmetric formulation with controls as the reference
is not recommended. Empirical results obtained from synthetic data are
presented, which confirm the benefit of properly accounting for potential
homogeneity under both matched and unmatched designs. We also present
preliminary results from the analysis a case-control study nested within the
EPIC cohort, where the objective is to identify metabolites associated with the
occurrence of subtypes of breast cancer.
| stat |
Sampling from manifold-restricted distributions using tangent bundle
projections | A common problem in Bayesian inference is the sampling of target probability
distributions at sufficient resolution and accuracy to estimate the probability
density, and to compute credible regions. Often by construction, many target
distributions can be expressed as some higher-dimensional closed-form
distribution with parametrically constrained variables, i.e., one that is
restricted to a smooth submanifold of Euclidean space. I propose a
derivative-based importance sampling framework for such distributions. A base
set of $n$ samples from the target distribution is used to map out the tangent
bundle of the manifold, and to seed $nm$ additional points that are projected
onto the tangent bundle and weighted appropriately. The method essentially acts
as an upsampling complement to any standard algorithm. It is designed for the
efficient production of approximate high-resolution histograms from
manifold-restricted Gaussian distributions, and can provide large computational
savings when sampling directly from the target distribution is expensive.
| stat |
On the Validity of Bayesian Neural Networks for Uncertainty Estimation | Deep neural networks (DNN) are versatile parametric models utilised
successfully in a diverse number of tasks and domains. However, they have
limitations---particularly from their lack of robustness and over-sensitivity
to out of distribution samples. Bayesian Neural Networks, due to their
formulation under the Bayesian framework, provide a principled approach to
building neural networks that address these limitations. This paper describes a
study that empirically evaluates and compares Bayesian Neural Networks to their
equivalent point estimate Deep Neural Networks to quantify the predictive
uncertainty induced by their parameters, as well as their performance in view
of this uncertainty. In this study, we evaluated and compared three point
estimate deep neural networks against comparable Bayesian neural network
alternatives using two well-known benchmark image classification datasets
(CIFAR-10 and SVHN).
| stat |
Online Multiclass Boosting with Bandit Feedback | We present online boosting algorithms for multiclass classification with
bandit feedback, where the learner only receives feedback about the correctness
of its prediction. We propose an unbiased estimate of the loss using a
randomized prediction, allowing the model to update its weak learners with
limited information. Using the unbiased estimate, we extend two full
information boosting algorithms (Jung et al., 2017) to the bandit setting. We
prove that the asymptotic error bounds of the bandit algorithms exactly match
their full information counterparts. The cost of restricted feedback is
reflected in the larger sample complexity. Experimental results also support
our theoretical findings, and performance of the proposed models is comparable
to that of an existing bandit boosting algorithm, which is limited to use
binary weak learners.
| stat |
Inference on Achieved Signal Noise Ratio | We describe a procedure to perform approximate inference on the achieved
signal-noise ratio of the Markowitz Portfolio under Gaussian i.i.d. returns.
The procedure relies on a statistic similar to the Sharpe Ratio Information
Criterion. Testing indicates the procedure is somewhat conservative, but
otherwise works well for reasonable values of sample and asset universe sizes.
We adapt the procedure to deal with generalizations of the portfolio
optimization problem.
| stat |
Monotone function estimation in the presence of extreme data coarsening:
Analysis of preeclampsia and birth weight in urban Uganda | This paper proposes a Bayesian hierarchical model to characterize the
relationship between birth weight and maternal pre-eclampsia across gestation
at a large maternity hospital in urban Uganda. Key scientific questions we
investigate include: 1) how pre-eclampsia compares to other maternal-fetal
covariates as a predictor of birth weight; and 2) whether the impact of
pre-eclampsia on birthweight varies across gestation. Our model addresses
several key statistical challenges: it correctly encodes the prior medical
knowledge that birth weight should vary smoothly and monotonically with
gestational age, yet it also avoids assumptions about functional form along
with assumptions about how birth weight varies with other covariates. Our model
also accounts for the fact that a high proportion (83%) of birth weights in our
data set are rounded to the nearest 100 grams. Such extreme data coarsening is
rare in maternity hospitals in high resource obstetrics settings but common for
data sets collected in low and middle-income countries (LMICs); this introduces
a substantial extra layer of uncertainty into the problem and is a major reason
why we adopt a Bayesian approach.
Our proposed non-parametric regression model, which we call Projective Smooth
BART (psBART), builds upon the highly successful Bayesian Additive Regression
Tree (BART) framework. This model captures complex nonlinear relationships and
interactions, induces smoothness and monotonicity in a single target covariate,
and provides a full posterior for uncertainty quantification. The results of
our analysis show that pre-eclampsia is a dominant predictor of birth weight in
this urban Ugandan setting, and therefore an important risk factor for
perinatal mortality.
| stat |
Data-Driven Model Set Design for Model Averaged Particle Filter | This paper is concerned with sequential state filtering in the presence of
nonlinearity, non-Gaussianity and model uncertainty. For this problem, the
Bayesian model averaged particle filter (BMAPF) is perhaps one of the most
efficient solutions. Major advances of BMAPF have been made, while it still
lacks a generic and practical approach to design the model set. This paper
fills in this gap by proposing a generic data-driven method for BMAPF model set
design. Unlike existent methods, the proposed solution does not require any
prior knowledge on the parameter value of the true model; it only assumes that
a small number of noisy observations are pre-obtained. The Bayesian
optimization (BO) method is adapted to search the model components, each of
which is associated with a specific segment of the pre-obtained dataset.The
average performance of these model components is guaranteed since each one's
parameter value is elaborately tuned via BO to maximize the marginal
likelihood. The diversity in the model components is also ensured, as different
components match the different segments of the pre-obtained dataset,
respectively. Computer simulations are used to demonstrate the effectiveness of
the proposed method.
| stat |
Jaccard/Tanimoto similarity test and estimation methods | Binary data are used in a broad area of biological sciences. Using binary
presence-absence data, we can evaluate species co-occurrences that help
elucidate relationships among organisms and environments. To summarize
similarity between occurrences of species, we routinely use the
Jaccard/Tanimoto coefficient, which is the ratio of their intersection to their
union. It is natural, then, to identify statistically significant
Jaccard/Tanimoto coefficients, which suggest non-random co-occurrences of
species. However, statistical hypothesis testing using this similarity
coefficient has been seldom used or studied.
We introduce a hypothesis test for similarity for biological presence-absence
data, using the Jaccard/Tanimoto coefficient. Several key improvements are
presented including unbiased estimation of expectation and centered
Jaccard/Tanimoto coefficients, that account for occurrence probabilities. We
derived the exact and asymptotic solutions and developed the bootstrap and
measurement concentration algorithms to compute statistical significance of
binary similarity. Comprehensive simulation studies demonstrate that our
proposed methods produce accurate p-values and false discovery rates. The
proposed estimation methods are orders of magnitude faster than the exact
solution. The proposed methods are implemented in an open source R package
called jaccard (https://cran.r-project.org/package=jaccard).
We introduce a suite of statistical methods for the Jaccard/Tanimoto
similarity coefficient, that enable straightforward incorporation of
probabilistic measures in analysis for species co-occurrences. Due to their
generality, the proposed methods and implementations are applicable to a wide
range of binary data arising from genomics, biochemistry, and other areas of
science.
| stat |
Persistence Bag-of-Words for Topological Data Analysis | Persistent homology (PH) is a rigorous mathematical theory that provides a
robust descriptor of data in the form of persistence diagrams (PDs). PDs
exhibit, however, complex structure and are difficult to integrate in today's
machine learning workflows. This paper introduces persistence bag-of-words: a
novel and stable vectorized representation of PDs that enables the seamless
integration with machine learning. Comprehensive experiments show that the new
representation achieves state-of-the-art performance and beyond in much less
time than alternative approaches.
| stat |
Learning with Fenchel-Young Losses | Over the past decades, numerous loss functions have been been proposed for a
variety of supervised learning tasks, including regression, classification,
ranking, and more generally structured prediction. Understanding the core
principles and theoretical properties underpinning these losses is key to
choose the right loss for the right problem, as well as to create new losses
which combine their strengths. In this paper, we introduce Fenchel-Young
losses, a generic way to construct a convex loss function for a regularized
prediction function. We provide an in-depth study of their properties in a very
broad setting, covering all the aforementioned supervised learning tasks, and
revealing new connections between sparsity, generalized entropies, and
separation margins. We show that Fenchel-Young losses unify many well-known
loss functions and allow to create useful new ones easily. Finally, we derive
efficient predictive and training algorithms, making Fenchel-Young losses
appealing both in theory and practice.
| stat |
A Statistical Approach to Assessing Neural Network Robustness | We present a new approach to assessing the robustness of neural networks
based on estimating the proportion of inputs for which a property is violated.
Specifically, we estimate the probability of the event that the property is
violated under an input model. Our approach critically varies from the formal
verification framework in that when the property can be violated, it provides
an informative notion of how robust the network is, rather than just the
conventional assertion that the network is not verifiable. Furthermore, it
provides an ability to scale to larger networks than formal verification
approaches. Though the framework still provides a formal guarantee of
satisfiability whenever it successfully finds one or more violations, these
advantages do come at the cost of only providing a statistical estimate of
unsatisfiability whenever no violation is found. Key to the practical success
of our approach is an adaptation of multi-level splitting, a Monte Carlo
approach for estimating the probability of rare events, to our statistical
robustness framework. We demonstrate that our approach is able to emulate
formal verification procedures on benchmark problems, while scaling to larger
networks and providing reliable additional information in the form of accurate
estimates of the violation probability.
| stat |
Machine learning for causal inference: on the use of cross-fit
estimators | Modern causal inference methods allow machine learning to be used to weaken
parametric modeling assumptions. However, the use of machine learning may
result in complications for inference. Doubly-robust cross-fit estimators have
been proposed to yield better statistical properties.
We conducted a simulation study to assess the performance of several
different estimators for the average causal effect (ACE). The data generating
mechanisms for the simulated treatment and outcome included log-transforms,
polynomial terms, and discontinuities. We compared singly-robust estimators
(g-computation, inverse probability weighting) and doubly-robust estimators
(augmented inverse probability weighting, targeted maximum likelihood
estimation). Nuisance functions were estimated with parametric models and
ensemble machine learning, separately. We further assessed doubly-robust
cross-fit estimators.
With correctly specified parametric models, all of the estimators were
unbiased and confidence intervals achieved nominal coverage. When used with
machine learning, the doubly-robust cross-fit estimators substantially
outperformed all of the other estimators in terms of bias, variance, and
confidence interval coverage.
Due to the difficulty of properly specifying parametric models in high
dimensional data, doubly-robust estimators with ensemble learning and
cross-fitting may be the preferred approach for estimation of the ACE in most
epidemiologic studies. However, these approaches may require larger sample
sizes to avoid finite-sample issues.
| stat |
Model Calibration via Distributionally Robust Optimization: On the NASA
Langley Uncertainty Quantification Challenge | We study a methodology to tackle the NASA Langley Uncertainty Quantification
Challenge, a model calibration problem under both aleatory and epistemic
uncertainties. Our methodology is based on an integration of robust
optimization, more specifically a recent line of research known as
distributionally robust optimization, and importance sampling in Monte Carlo
simulation. The main computation machinery in this integrated methodology
amounts to solving sampled linear programs. We present theoretical statistical
guarantees of our approach via connections to nonparametric hypothesis testing,
and numerical performances including parameter calibration and downstream
decision and risk evaluation tasks.
| stat |
Minimax Policy for Heavy-tailed Bandits | We study the stochastic Multi-Armed Bandit (MAB) problem under worst-case
regret and heavy-tailed reward distribution. We modify the minimax policy MOSS
for the sub-Gaussian reward distribution by using saturated empirical mean to
design a new algorithm called Robust MOSS. We show that if the moment of order
$1+\epsilon$ for the reward distribution exists, then the refined strategy has
a worst-case regret matching the lower bound while maintaining a
distribution-dependent logarithm regret.
| stat |
Massively scalable Sinkhorn distances via the Nystr\"om method | The Sinkhorn "distance", a variant of the Wasserstein distance with entropic
regularization, is an increasingly popular tool in machine learning and
statistical inference. However, the time and memory requirements of standard
algorithms for computing this distance grow quadratically with the size of the
data, making them prohibitively expensive on massive data sets. In this work,
we show that this challenge is surprisingly easy to circumvent: combining two
simple techniques---the Nystr\"om method and Sinkhorn scaling---provably yields
an accurate approximation of the Sinkhorn distance with significantly lower
time and memory requirements than other approaches. We prove our results via
new, explicit analyses of the Nystr\"om method and of the stability properties
of Sinkhorn scaling. We validate our claims experimentally by showing that our
approach easily computes Sinkhorn distances on data sets hundreds of times
larger than can be handled by other techniques.
| stat |
Nonparametric Estimation of Population Average Dose-Response Curves
using Entropy Balancing Weights for Continuous Exposures | Weighted estimators are commonly used for estimating exposure effects in
observational settings to establish causal relations. These estimators have a
long history of development when the exposure of interest is binary and where
the weights are typically functions of an estimated propensity score. Recent
developments in optimization-based estimators for constructing weights in
binary exposure settings, such as those based on entropy balancing, have shown
more promise in estimating treatment effects than those methods that focus on
the direct estimation of the propensity score using likelihood-based methods.
This paper explores recent developments of entropy balancing methods to
continuous exposure settings and the estimation of population dose-response
curves using nonparametric estimation combined with entropy balancing weights,
focusing on factors that would be important to applied researchers in medical
or health services research. The methods developed here are applied to data
from a study assessing the effect of non-randomized components of an
evidence-based substance use treatment program on emotional and substance use
clinical outcomes.
| stat |
Bayesian Ordinal Quantile Regression with a Partially Collapsed Gibbs
Sampler | Unlike standard linear regression, quantile regression captures the
relationship between covariates and the conditional response distribution as a
whole, rather than only the relationship between covariates and the expected
value of the conditional response. However, while there are well-established
quantile regression methods for continuous variables and some forms of discrete
data, there is no widely accepted method for ordinal variables, despite their
importance in many medical contexts. In this work, we describe two existing
ordinal quantile regression methods and demonstrate their weaknesses. We then
propose a new method, Bayesian ordinal quantile regression with a partially
collapsed Gibbs sampler (BORPS). We show superior results using BORPS versus
existing methods on an extensive set of simulations. We further illustrate the
benefits of our method by applying BORPS to the Fragile Families and Child
Wellbeing Study data to tease apart associations with early puberty among both
genders. Software is available at: GitHub.com/igrabski/borps.
| stat |
Quantifying Uncertainty in Infectious Disease Mechanistic Models | This primer describes the statistical uncertainty in mechanistic models and
provides R code to quantify it. We begin with an overview of mechanistic models
for infectious disease, and then describe the sources of statistical
uncertainty in the context of a case study on SARS-CoV-2. We describe the
statistical uncertainty as belonging to three categories: data uncertainty,
stochastic uncertainty, and structural uncertainty. We demonstrate how to
account for each of these via statistical uncertainty measures and sensitivity
analyses broadly, as well as in a specific case study on estimating the basic
reproductive number, $R_0$, for SARS-CoV-2.
| stat |
State space models for non-stationary intermittently coupled systems: an
application to the North Atlantic Oscillation | We develop Bayesian state space methods for modelling changes to the mean
level or temporal correlation structure of an observed time series due to
intermittent coupling with an unobserved process. Novel intervention methods
are proposed to model the effect of repeated coupling as a single dynamic
process. Latent time-varying autoregressive components are developed to model
changes in the temporal correlation structure. Efficient filtering and
smoothing methods are derived for the resulting class of models. We propose
methods for quantifying the component of variance attributable to an unobserved
process, the effect during individual coupling events, and the potential for
skilful forecasts.
The proposed methodology is applied to the study of winter-time variability
in the dominant pattern of climate variation in the northern hemisphere, the
North Atlantic Oscillation. Around 70% of the inter-annual variance in the
winter (Dec-Jan-Feb) mean level is attributable to an unobserved process.
Skilful forecasts for winter (Dec-Jan-Feb) mean are possible from the beginning
of December.
| stat |
Semiparametric regression and risk prediction with competing risks data
under missing cause of failure | The cause of failure in cohort studies that involve competing risks is
frequently incompletely observed. To address this, several methods have been
proposed for the semiparametric proportional cause-specific hazards model under
a missing at random assumption. However, these proposals provide inference for
the regression coefficients only, and do not consider the infinite dimensional
parameters, such as the covariate-specific cumulative incidence function.
Nevertheless, the latter quantity is essential for risk prediction in modern
medicine. In this paper we propose a unified framework for inference about both
the regression coefficients of the proportional cause-specific hazards model
and the covariate-specific cumulative incidence functions under missing at
random cause of failure. Our approach is based on a novel computationally
efficient maximum pseudo-partial-likelihood estimation method for the
semiparametric proportional cause-specific hazards model. Using modern
empirical process theory we derive the asymptotic properties of the proposed
estimators for the regression coefficients and the covariate-specific
cumulative incidence functions, and provide methodology for constructing
simultaneous confidence bands for the latter. Simulation studies show that our
estimators perform well even in the presence of a large fraction of missing
cause of failures, and that the regression coefficient estimator can be
substantially more efficient compared to the previously proposed augmented
inverse probability weighting estimator. The method is applied using data from
an HIV cohort study and a bladder cancer clinical trial.
| stat |
Uncovering differential equations from data with hidden variables | SINDy is a method for learning system of differential equations from data by
solving a sparse linear regression optimization problem [Brunton et al., 2016].
In this article, we propose an extension of the SINDy method that learns
systems of differential equations in cases where some of the variables are not
observed. Our extension is based on regressing a higher order time derivative
of a target variable onto a dictionary of functions that includes lower order
time derivatives of the target variable. We evaluate our method by measuring
the prediction accuracy of the learned dynamical systems on synthetic data and
on a real data-set of temperature time series provided by the R\'eseau de
Transport d'\'Electricit\'e (RTE). Our method provides high quality short-term
forecasts and it is orders of magnitude faster than competing methods for
learning differential equations with latent variables.
| stat |
Sparse Regression: Scalable algorithms and empirical performance | In this paper, we review state-of-the-art methods for feature selection in
statistics with an application-oriented eye. Indeed, sparsity is a valuable
property and the profusion of research on the topic might have provided little
guidance to practitioners. We demonstrate empirically how noise and correlation
impact both the accuracy - the number of correct features selected - and the
false detection - the number of incorrect features selected - for five methods:
the cardinality-constrained formulation, its Boolean relaxation, $\ell_1$
regularization and two methods with non-convex penalties. A cogent feature
selection method is expected to exhibit a two-fold convergence, namely the
accuracy and false detection rate should converge to $1$ and $0$ respectively,
as the sample size increases. As a result, proper method should recover all and
nothing but true features. Empirically, the integer optimization formulation
and its Boolean relaxation are the closest to exhibit this two properties
consistently in various regimes of noise and correlation. In addition, apart
from the discrete optimization approach which requires a substantial, yet often
affordable, computational time, all methods terminate in times comparable with
the \verb|glmnet| package for Lasso. We released code for methods that were not
publicly implemented. Jointly considered, accuracy, false detection and
computational time provide a comprehensive assessment of each feature selection
method and shed light on alternatives to the Lasso-regularization which are not
as popular in practice yet.
| stat |
Accounting for total variation and robustness in profiling health care
providers | Monitoring outcomes of health care providers, such as patient deaths,
hospitalizations and hospital readmissions, helps in assessing the quality of
health care. We consider a large database on patients being treated at dialysis
facilities in the United States, and the problem of identifying facilities with
outcomes that are better than or worse than expected. Analyses of such data
have been commonly based on random or fixed facility effects, which have
shortcomings that can lead to unfair assessments. A primary issue is that they
do not appropriately account for variation between providers that is outside
the providers' control due, for example, to unobserved patient characteristics
that vary between providers. In this article, we propose a smoothed empirical
null approach that accounts for the total variation and adapts to different
provider sizes. The linear model provides an illustration that extends easily
to other nonlinear models for survival or binary outcomes, for example. The
empirical null method is generalized to allow for some variation being due to
quality of care. These methods are examined with numerical simulations and
applied to the monitoring of survival in the dialysis facility data.
| stat |
Global Fitting of the Response Surface via Estimating Multiple Contours
of a Simulator | Computer simulators are nowadays widely used to understand complex physical
systems in many areas such as aerospace, renewable energy, climate modeling,
and manufacturing. One fundamental issue in the study of computer simulators is
known as experimental design, that is, how to select the input settings where
the computer simulator is run and the corresponding response is collected.
Extra care should be taken in the selection process because computer simulators
can be computationally expensive to run. The selection shall acknowledge and
achieve the goal of the analysis. This article focuses on the goal of producing
more accurate prediction which is important for risk assessment and decision
making. We propose two new methods of design approaches that sequentially
select input settings to achieve this goal. The approaches make novel
applications of simultaneous and sequential contour estimations. Numerical
examples are employed to demonstrate the effectiveness of the proposed
approaches.
| stat |
Personalized Decision Making for Biopsies in Prostate Cancer Active
Surveillance Programs | Background: Low-risk prostate cancer patients enrolled in active surveillance
programs commonly undergo biopsies for examination of cancer progression.
Biopsies are conducted as per a fixed and frequent schedule (e.g., annual
biopsies). Since biopsies are burdensome, patients do not always comply with
the schedule, which increases the risk of delayed detection of cancer
progression.
Objective: Our aim is to better balance the number of biopsies (burden) and
the delay in detection of cancer progression (less is beneficial), by
personalizing the decision of conducting biopsies.
Data Sources: We use patient data of the world's largest active surveillance
program (PRIAS). It enrolled 5270 patients, had 866 cancer progressions, and an
average of nine prostate-specific antigen (PSA) and five digital rectal
examination (DRE) measurements per patient.
Methods: Using joint models for time-to-event and longitudinal data, we model
the historical DRE and PSA measurements, and biopsy results of a patient at
each follow-up visit. This results in a visit and patient-specific cumulative
risk of cancer progression. If this risk is above a certain threshold, we
schedule a biopsy. We compare this personalized approach with the currently
practiced biopsy schedules via an extensive and realistic simulation study,
based on a replica of the patients from the PRIAS program.
Results: The personalized approach saved a median of six biopsies (median: 4,
IQR: 2-5), compared to the annual schedule (median: 10, IQR: 3-10). However,
the delay in detection of progression (years) is similar for the personalized
(median: 0.7, IQR: 0.3-1.0) and the annual schedule (median: 0.5, IQR:
0.3-0.8).
Conclusions: We conclude that personalized schedules provide substantially
better balance in the number of biopsies per detected progression for men with
low-risk prostate cancer.
| stat |
GO Hessian for Expectation-Based Objectives | An unbiased low-variance gradient estimator, termed GO gradient, was proposed
recently for expectation-based objectives
$\mathbb{E}_{q_{\boldsymbol{\gamma}}(\boldsymbol{y})} [f(\boldsymbol{y})]$,
where the random variable (RV) $\boldsymbol{y}$ may be drawn from a stochastic
computation graph with continuous (non-reparameterizable) internal nodes and
continuous/discrete leaves. Upgrading the GO gradient, we present for
$\mathbb{E}_{q_{\boldsymbol{\boldsymbol{\gamma}}}(\boldsymbol{y})}
[f(\boldsymbol{y})]$ an unbiased low-variance Hessian estimator, named GO
Hessian. Considering practical implementation, we reveal that GO Hessian is
easy-to-use with auto-differentiation and Hessian-vector products, enabling
efficient cheap exploitation of curvature information over stochastic
computation graphs. As representative examples, we present the GO Hessian for
non-reparameterizable gamma and negative binomial RVs/nodes. Based on the GO
Hessian, we design a new second-order method for
$\mathbb{E}_{q_{\boldsymbol{\boldsymbol{\gamma}}}(\boldsymbol{y})}
[f(\boldsymbol{y})]$, with rigorous experiments conducted to verify its
effectiveness and efficiency.
| stat |
Gradient flow encoding with distance optimization adaptive step size | The autoencoder model uses an encoder to map data samples to a lower
dimensional latent space and then a decoder to map the latent space
representations back to the data space. Implicitly, it relies on the encoder to
approximate the inverse of the decoder network, so that samples can be mapped
to and back from the latent space faithfully. This approximation may lead to
sub-optimal latent space representations. In this work, we investigate a
decoder-only method that uses gradient flow to encode data samples in the
latent space. The gradient flow is defined based on a given decoder and aims to
find the optimal latent space representation for any given sample through
optimisation, eliminating the need of an approximate inversion through an
encoder. Implementing gradient flow through ordinary differential equations
(ODE), we leverage the adjoint method to train a given decoder. We further show
empirically that the costly integrals in the adjoint method may not be entirely
necessary. Additionally, we propose a $2^{nd}$ order ODE variant to the method,
which approximates Nesterov's accelerated gradient descent, with faster
convergence per iteration. Commonly used ODE solvers can be quite sensitive to
the integration step-size depending on the stiffness of the ODE. To overcome
the sensitivity for gradient flow encoding, we use an adaptive solver that
prioritises minimising loss at each integration step. We assess the proposed
method in comparison to the autoencoding model. In our experiments, GFE showed
a much higher data-efficiency than the autoencoding model, which can be crucial
for data scarce applications.
| stat |
Bessel regression model: Robustness to analyze bounded data | Beta regression has been extensively used by statisticians and practitioners
to model bounded continuous data and there is no strong and similar competitor
having its main features. A class of normalized inverse-Gaussian (N-IG) process
was introduced in the literature, being explored in the Bayesian context as a
powerful alternative to the Dirichlet process. Until this moment, no attention
has been paid for the univariate N-IG distribution in the classical inference.
In this paper, we propose the bessel regression based on the univariate N-IG
distribution, which is a robust alternative to the beta model. This robustness
is illustrated through simulated and real data applications. The estimation of
the parameters is done through an Expectation-Maximization algorithm and the
paper discusses how to perform inference. A useful and practical discrimination
procedure is proposed for model selection between bessel and beta regressions.
Monte Carlo simulation results are presented to verify the finite-sample
behavior of the EM-based estimators and the discrimination procedure. Further,
the performances of the regressions are evaluated under misspecification, which
is a critical point showing the robustness of the proposed model. Finally,
three empirical illustrations are explored to confront results from bessel and
beta regressions.
| stat |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.