title
stringlengths 6
244
| abstract
stringlengths 19
5.09k
| label
stringclasses 10
values |
|---|---|---|
Bayesian analysis of multifidelity computer models with local features
and non-nested experimental designs: Application to the WRF model | We propose a multi-fidelity Bayesian emulator for the analysis of the Weather
Research and Forecasting (WRF) model when the available simulations are not
generated based on hierarchically nested experimental design. The proposed
procedure, called Augmented Bayesian Treed Co-Kriging, extends the scope of
co-kriging in two major ways. We introduce a binary treed partition latent
process in the multifidelity setting to account for non-stationary and
potential discontinuities in the model outputs at different fidelity levels.
Moreover, we introduce an efficient imputation mechanism which allows the
practical implementation of co-kriging when the experimental design is
non-hierarchically nested by enabling the specification of semi-conjugate
priors. Our imputation strategy allows the design of an efficient RJ-MCMC
implementation that involves collapsed blocks and direct simulation from
conditional distributions. We develop the Monte Carlo recursive emulator which
provides a Monte Carlo proxy for the full predictive distribution of the model
output at each fidelity level, in a computationally feasible manner. The
performance of our method is demonstrated on a benchmark example, and compared
against existing methods. The proposed method is used for the analysis of a
large-scale climate modeling application which involves the WRF model.
| stat |
Adding flexibility to clinical trial designs: an example-based guide to
the practical use of adaptive designs | Adaptive designs for clinical trials permit alterations to a study in
response to accumulating data in order to make trials more flexible, ethical
and efficient. These benefits are achieved while preserving the integrity and
validity of the trial, through the pre-specification and proper adjustment for
the possible alterations during the course of the trial. Despite much research
in the statistical literature highlighting the potential advantages of adaptive
designs over traditional fixed designs, the uptake of such methods in clinical
research has been slow. One major reason for this is that different adaptations
to trial designs, as well as their advantages and limitations, remain
unfamiliar to large parts of the clinical community. The aim of this paper is
to clarify where adaptive designs can be used to address specific questions of
scientific interest; we introduce the main features of adaptive designs and
commonly used terminology, highlighting their utility and pitfalls, and
illustrate their use through case studies of adaptive trials ranging from
early-phase dose escalation to confirmatory Phase III studies.
| stat |
Outcome model free causal inference with ultra-high dimensional
covariates | Causal inference has been increasingly reliant on observational studies with
rich covariate information. To build tractable causal models, including the
propensity score models, it is imperative to first extract important features
from high dimensional data. Unlike the familiar task of variable selection for
prediction modeling, our feature selection procedure aims to control for
confounding while maintaining efficiency in the resulting causal effect
estimate. Previous empirical studies imply that one should aim to include all
predictors of the outcome, rather than the treatment, in the propensity score
model. In this paper, we formalize this intuition through rigorous proofs, and
propose the causal ball screening for selecting these variables from modern
ultra-high dimensional data sets. A distinctive feature of our proposal is that
we do not require any modeling on the outcome regression, thus providing
robustness against misspecification of the functional form or violation of
smoothness conditions. Our theoretical analyses show that the proposed
procedure enjoys a number of oracle properties including model selection
consistency, normality and efficiency. Synthetic and real data analyses show
that our proposal performs favorably with existing methods in a range of
realistic settings.
| stat |
False Discovery Rate Control via Data Splitting | Selecting relevant features associated with a given response variable is an
important issue in many scientific fields. Quantifying quality and uncertainty
of a selection result via false discovery rate (FDR) control has been of recent
interest. This paper introduces a way of using data-splitting strategies to
asymptotically control the FDR while maintaining a high power. For each
feature, the method constructs a test statistic by estimating two independent
regression coefficients via data splitting. FDR control is achieved by taking
advantage of the statistic's property that, for any null feature, its sampling
distribution is symmetric about zero. Furthermore, we propose Multiple Data
Splitting (MDS) to stabilize the selection result and boost the power.
Interestingly and surprisingly, with the FDR still under control, MDS not only
helps overcome the power loss caused by sample splitting, but also results in a
lower variance of the false discovery proportion (FDP) compared with all other
methods in consideration. We prove that the proposed data-splitting methods can
asymptotically control the FDR at any designated level for linear and Gaussian
graphical models in both low and high dimensions. Through intensive simulation
studies and a real-data application, we show that the proposed methods are
robust to the unknown distribution of features, easy to implement and
computationally efficient, and are often the most powerful ones amongst
competitors especially when the signals are weak and the correlations or
partial correlations are high among features.
| stat |
Large Sample Asymptotics of the Pseudo-Marginal Method | The pseudo-marginal algorithm is a variant of the Metropolis--Hastings
algorithm which samples asymptotically from a probability distribution when it
is only possible to estimate unbiasedly an unnormalized version of its density.
Practically, one has to trade-off the computational resources used to obtain
this estimator against the asymptotic variances of the ergodic averages
obtained by the pseudo-marginal algorithm. Recent works optimizing this
trade-off rely on some strong assumptions which can cast doubts over their
practical relevance. In particular, they all assume that the distribution of
the difference between the log-density and its estimate is independent of the
parameter value at which it is evaluated. Under regularity conditions we show
here that, as the number of data points tends to infinity, a space-rescaled
version of the pseudo-marginal chain converges weakly towards another
pseudo-marginal chain for which this assumption indeed holds. A study of this
limiting chain allows us to provide parameter dimension-dependent guidelines on
how to optimally scale a normal random walk proposal and the number of Monte
Carlo samples for the pseudo-marginal method in the large-sample regime. This
complements and validates currently available results.
| stat |
Effects of West Coast forest fire emissions on atmospheric environment:
A coupled satellite and ground-based assessment | Forest fires have a profound impact on the atmospheric environment and air
quality across the ecosystems. The recent west coast forest fire in the United
States of America (USA) has broken all the past records and caused severe
environmental and public health burdens. As of middle September, nearly 6
million acres forest area were burned, and more than 25 casualties were
reported so far. In this study, both satellite and in-situ air pollution data
were utilized to examine the effects of this unprecedented wildfire on the
atmospheric environment. The spatiotemporal concentrations of total six air
pollutants, i.e. carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide
(SO2), ozone (O3), particulate matter (PM2.5 and PM10), and aerosol index (AI),
were measured for the periods of 15 August to 15 September for 2020 (fire year)
and 2019 (reference year). The in-situ data-led measurements show that the
highest increases in CO (ppm), PM2.5, and PM10 concentrations ({\mu}g/m3) were
clustered around the west coastal fire-prone states, during the 15 August - 15
September period. The average CO concentration (ppm) was increased most
significantly in Oregon (1147.10), followed by Washington (812.76), and
California (13.17). Meanwhile, the concentration ({\mu}g/m3) in particulate
matter (both PM2.5 and PM10), was increased in all three states affected
severely by wildfires. Changes (positive) in both PM2.5 and PM10 were measured
highest in Washington (45.83 and 88.47 for PM2.5 and PM10), followed by Oregon
(41.99 and 62.75 for PM2.5 and PM10), and California (31.27 and 35.04 for PM2.5
and PM10). The average level of exposure to CO, PM2.5, and PM10 was also
measured for all the three fire-prone states. The results of the exposure
assessment revealed a strong tradeoff association between wildland fire and
local/regional air quality standard.
| stat |
SEIRD Model for Qatar Covid-19 Outbreak: A Case Study | The Covid-19 outbreak of 2020 has required many governments to develop
mathematical-statistical models of the outbreak for policy and planning
purposes. This work provides a tutorial on building a compartmental model using
Susceptibles, Exposed, Infected, Recovered and Deaths status through time. A
Bayesian Framework is utilized to perform both parameter estimation and
predictions. This model uses interventions to quantify the impact of various
government attempts to slow the spread of the virus. Predictions are also made
to determine when the peak Active Infections will occur.
| stat |
Asymptotic learning curves of kernel methods: empirical data v.s.
Teacher-Student paradigm | How many training data are needed to learn a supervised task? It is often
observed that the generalization error decreases as $n^{-\beta}$ where $n$ is
the number of training examples and $\beta$ an exponent that depends on both
data and algorithm. In this work we measure $\beta$ when applying kernel
methods to real datasets. For MNIST we find $\beta\approx 0.4$ and for CIFAR10
$\beta\approx 0.1$, for both regression and classification tasks, and for
Gaussian or Laplace kernels. To rationalize the existence of non-trivial
exponents that can be independent of the specific kernel used, we study the
Teacher-Student framework for kernels. In this scheme, a Teacher generates data
according to a Gaussian random field, and a Student learns them via kernel
regression. With a simplifying assumption -- namely that the data are sampled
from a regular lattice -- we derive analytically $\beta$ for translation
invariant kernels, using previous results from the kriging literature. Provided
that the Student is not too sensitive to high frequencies, $\beta$ depends only
on the smoothness and dimension of the training data. We confirm numerically
that these predictions hold when the training points are sampled at random on a
hypersphere. Overall, the test error is found to be controlled by the magnitude
of the projection of the true function on the kernel eigenvectors whose rank is
larger than $n$. Using this idea we predict relate the exponent $\beta$ to an
exponent $a$ describing how the coefficients of the true function in the
eigenbasis of the kernel decay with rank. We extract $a$ from real data by
performing kernel PCA, leading to $\beta\approx0.36$ for MNIST and
$\beta\approx0.07$ for CIFAR10, in good agreement with observations. We argue
that these rather large exponents are possible due to the small effective
dimension of the data.
| stat |
Regret Analysis of Bandit Problems with Causal Background Knowledge | We study how to learn optimal interventions sequentially given causal
information represented as a causal graph along with associated conditional
distributions. Causal modeling is useful in real world problems like online
advertisement where complex causal mechanisms underlie the relationship between
interventions and outcomes. We propose two algorithms, causal upper confidence
bound (C-UCB) and causal Thompson Sampling (C-TS), that enjoy improved
cumulative regret bounds compared with algorithms that do not use causal
information. We thus resolve an open problem posed by
\cite{lattimore2016causal}. Further, we extend C-UCB and C-TS to the linear
bandit setting and propose causal linear UCB (CL-UCB) and causal linear TS
(CL-TS) algorithms. These algorithms enjoy a cumulative regret bound that only
scales with the feature dimension. Our experiments show the benefit of using
causal information. For example, we observe that even with a few hundreds of
iterations, the regret of causal algorithms is less than that of standard
algorithms by a factor of three. We also show that under certain causal
structures, our algorithms scale better than the standard bandit algorithms as
the number of interventions increases.
| stat |
A Mortality Model for Multi-populations: A Semi-Parametric Approach | Mortality is different across countries, states and regions. Several
empirical research works however reveal that mortality trends exhibit a common
pattern and show similar structures across populations. The key element in
analyzing mortality rate is a time-varying indicator curve. Our main interest
lies in validating the existence of the common trends among these curves, the
similar gender differences and their variability in location among the curves
at the national level. Motivated by the empirical findings, we make the study
of estimating and forecasting mortality rates based on a semi-parametric
approach, which is applied to multiple curves with the shape-related nonlinear
variation. This approach allows us to capture the common features contained in
the curve functions and meanwhile provides the possibility to characterize the
nonlinear variation via a few deviation parameters. These parameters carry an
instructive summary of the time-varying curve functions and can be further used
to make a suggestive forecast analysis for countries with barren data sets. In
this research the model is illustrated with mortality rates of Japan and China,
and extended to incorporate more countries.
| stat |
Multimodal Data Integration via Mediation Analysis with High-Dimensional
Exposures and Mediators | Motivated by an imaging proteomics study for Alzheimer's disease (AD), in
this article, we propose a mediation analysis approach with high-dimensional
exposures and high-dimensional mediators to integrate data collected from
multiple platforms. The proposed method combines principal component analysis
with penalized least squares estimation for a set of linear structural equation
models. The former reduces the dimensionality and produces uncorrelated linear
combinations of the exposure variables, whereas the latter achieves
simultaneous path selection and effect estimation while allowing the mediators
to be correlated. Applying the method to the AD data identifies numerous
interesting protein peptides, brain regions, and protein-structure-memory
paths, which are in accordance with and also supplement existing findings of AD
research. Additional simulations further demonstrate the effective empirical
performance of the method.
| stat |
An Efficient Algorithm for Elastic I-optimal Design of Generalized
Linear Models | The generalized linear models (GLMs) are widely used in statistical analysis
and the related design issues are undoubtedly challenging. The state-of-the-art
works mostly apply to design criteria on the estimates of regression
coefficients. The prediction accuracy is usually critical in modern decision
making and artificial intelligence applications. It is of importance to study
optimal designs from the prediction aspects for generalized linear models. In
this work, we consider the Elastic I-optimality as a prediction-oriented design
criterion for generalized linear models, and develop efficient algorithms for
such $\text{EI}$-optimal designs. By investigating theoretical properties for
the optimal weights of any set of design points and extending the general
equivalence theorem to the $\text{EI}$-optimality for GLMs, the proposed
efficient algorithm adequately combines the Fedorov-Wynn algorithm and
multiplicative algorithm. It achieves great computational efficiency with
guaranteed convergence. Numerical examples are conducted to evaluate the
feasibility and computational efficiency of the proposed algorithm.
| stat |
Rerandomization in stratified randomized experiments | Stratification and rerandomization are two well-known methods used in
randomized experiments for balancing the baseline covariates. Renowned scholars
in experimental design have recommended combining these two methods; however,
limited studies have addressed the statistical properties of this combination.
This paper proposes two rerandomization methods to be used in stratified
randomized experiments, based on the overall and stratum-specific Mahalanobis
distances. The first method is applicable for nearly arbitrary numbers of
strata, strata sizes, and stratum-specific proportions of the treated units.
The second method, which is generally more efficient than the first method, is
suitable for situations in which the number of strata is fixed with their sizes
tending to infinity. Under the randomization inference framework, we obtain
these methods' asymptotic distributions and the formulas of variance reduction
when compared to stratified randomization. Our analysis does not require any
modeling assumption regarding the potential outcomes. Moreover, we provide
asymptotically conservative variance estimators and confidence intervals for
the average treatment effect. The advantages of the proposed methods are
exhibited through an extensive simulation study and a real-data example.
| stat |
Scalable high-resolution forecasting of sparse spatiotemporal events
with kernel methods: a winning solution to the NIJ "Real-Time Crime
Forecasting Challenge" | We propose a generic spatiotemporal event forecasting method, which we
developed for the National Institute of Justice's (NIJ) Real-Time Crime
Forecasting Challenge. Our method is a spatiotemporal forecasting model
combining scalable randomized Reproducing Kernel Hilbert Space (RKHS) methods
for approximating Gaussian processes with autoregressive smoothing kernels in a
regularized supervised learning framework. While the smoothing kernels capture
the two main approaches in current use in the field of crime forecasting,
kernel density estimation (KDE) and self-exciting point process (SEPP) models,
the RKHS component of the model can be understood as an approximation to the
popular log-Gaussian Cox Process model. For inference, we discretize the
spatiotemporal point pattern and learn a log-intensity function using the
Poisson likelihood and highly efficient gradient-based optimization methods.
Model hyperparameters including quality of RKHS approximation, spatial and
temporal kernel lengthscales, number of autoregressive lags, bandwidths for
smoothing kernels, as well as cell shape, size, and rotation, were learned
using crossvalidation. Resulting predictions significantly exceeded baseline
KDE estimates and SEPP models for sparse events.
| stat |
An Ensemble Method for Interval-Censored Time-to-Event Data | Interval-censored data analysis is important in biomedical statistics for any
type of time-to-event response where the time of response is not known exactly,
but rather only known to occur between two assessment times. Many clinical
trials and longitudinal studies generate interval-censored data; one common
example occurs in medical studies that entail periodic follow-up. In this paper
we propose a survival forest method for interval-censored data based on the
conditional inference framework. We describe how this framework can be adapted
to the situation of interval-censored data. We show that the tuning parameters
have a non-negligible effect on the survival forest performance and guidance is
provided on how to tune the parameters in a data-dependent way to improve the
overall performance of the method. Using Monte Carlo simulations we find that
the proposed survival forest is at least as effective as a survival tree method
when the underlying model has a tree structure, performs similarly to an
interval-censored Cox proportional hazards model fit when the true relationship
is linear, and outperforms the survival tree method and Cox model when the true
relationship is nonlinear. We illustrate the application of the method on a
tooth emergence data set.
| stat |
NeRF-VAE: A Geometry Aware 3D Scene Generative Model | We propose NeRF-VAE, a 3D scene generative model that incorporates geometric
structure via NeRF and differentiable volume rendering. In contrast to NeRF,
our model takes into account shared structure across scenes, and is able to
infer the structure of a novel scene -- without the need to re-train -- using
amortized inference. NeRF-VAE's explicit 3D rendering process further contrasts
previous generative models with convolution-based rendering which lacks
geometric structure. Our model is a VAE that learns a distribution over
radiance fields by conditioning them on a latent scene representation. We show
that, once trained, NeRF-VAE is able to infer and render
geometrically-consistent scenes from previously unseen 3D environments using
very few input images. We further demonstrate that NeRF-VAE generalizes well to
out-of-distribution cameras, while convolutional models do not. Finally, we
introduce and study an attention-based conditioning mechanism of NeRF-VAE's
decoder, which improves model performance.
| stat |
Continual Learning in Practice | This paper describes a reference architecture for self-maintaining systems
that can learn continually, as data arrives. In environments where data
evolves, we need architectures that manage Machine Learning (ML) models in
production, adapt to shifting data distributions, cope with outliers, retrain
when necessary, and adapt to new tasks. This represents continual AutoML or
Automatically Adaptive Machine Learning. We describe the challenges and
proposes a reference architecture.
| stat |
Designing experiments for estimating an appropriate outlet size for a
silo type problem | The problem of jam formation during the discharge by gravity of granular
material through a two-dimensional silo has a number of practical applications.
In many problems the estimation of the minimum outlet size which guarantees
that the time to the next jamming event is long enough is crucial. Assuming
that the time is modeled by an exponential distribution with two unknown
parameters, this goal translates to the optimal estimation of a non-linear
transformation of the parameters. We obtain $c$-optimum experimental designs
with that purpose, applying the graphic Elfving method. Since the optimal
designs depend on the nominal values of the parameters, a sensitivity study is
additionally provided. Finally, a simulation study checks the performance of
the approximations made, first with the Fisher Information matrix, then with
the linearization of the function to be estimated. The results are useful for
experimenting in a laboratory and translating then the results to a larger
scenario. Apart from the application a general methodology is developed in the
paper for the problem of precise estimation of a one-dimensional parametric
transformation in a non-linear model.
| stat |
Boosted nonparametric hazards with time-dependent covariates | Given functional data from a survival process with time-dependent covariates,
we derive a smooth convex representation for its nonparametric log-likelihood
functional and obtain its functional gradient. From this we devise a generic
gradient boosting procedure for estimating the hazard function
nonparametrically. An illustrative implementation of the procedure using
regression trees is described to show how to recover the unknown hazard. We
show that the generic estimator is consistent if the model is correctly
specified; alternatively an oracle inequality can be demonstrated for
tree-based models. To avoid overfitting, boosting employs several
regularization devices. One of them is step-size restriction, but the rationale
for this is somewhat mysterious from the viewpoint of consistency. Our work
brings some clarity to this issue by revealing that step-size restriction is a
mechanism for preventing the curvature of the risk from derailing convergence.
| stat |
Electoral David vs Goliath: How does the Spatial Concentration of
Electors affect District-based Elections? | Many democratic countries use district-based elections where there is a
"seat" for each district in the governing body. In each district, the party
whose candidate gets the maximum number of votes wins the corresponding seat.
The result of the election is decided based on the number of seats won by the
different parties. The electors (voters) can cast their votes only in the
district of their residence. Thus, locations of the electors and boundaries of
the districts may severely affect the election result even if the proportion of
popular support (number of electors) of different parties remains unchanged.
This has led to significant amount of research on whether the districts may be
redrawn or electors may be moved to maximize seats for a particular party. In
this paper, we frame the spatial distribution of electors in a probabilistic
setting, and explore different models to capture the intra-district
polarization of electors in favour of a party, or the spatial concentration of
supporters of different parties. Our models are inspired by elections in India,
where supporters of different parties tend to be concentrated in certain
districts. We show with extensive simulations that our model can capture
different statistical properties of real elections held in India. We frame
parameter estimation problems to fit our models to the observed election
results. Since analytical calculation of the likelihood functions are
infeasible for our complex models, we use Likelihood-free Inference methods
under the Approximate Bayesian Computation framework. Since this approach is
highly time-consuming, we explore how supervised regression using Logistic
Regression or Deep Neural Networks can be used to speed it up. We also explore
how the election results can change by varying the spatial distributions of the
voters, even when the proportions of popular support of the parties remain
constant.
| stat |
Kernel Partial Correlation Coefficient -- a Measure of Conditional
Dependence | In this paper we propose and study a class of simple, nonparametric, yet
interpretable measures of conditional dependence between two random variables
$Y$ and $Z$ given a third variable $X$, all taking values in general
topological spaces. The population version of any of these measures captures
the strength of conditional dependence and it is 0 if and only if $Y$ and $Z$
are conditionally independent given $X$, and 1 if and only if $Y$ is a
measurable function of $Z$ and $X$. Thus, our measure -- which we call kernel
partial correlation (KPC) coefficient -- can be thought of as a nonparametric
generalization of the partial correlation coefficient that possesses the above
properties when $(X,Y,Z)$ is jointly normal. We describe two consistent methods
of estimating KPC. Our first method utilizes the general framework of geometric
graphs, including $K$-nearest neighbor graphs and minimum spanning trees. A
sub-class of these estimators can be computed in near linear time and converges
at a rate that automatically adapts to the intrinsic dimension(s) of the
underlying distribution(s). Our second strategy involves direct estimation of
conditional mean embeddings using cross-covariance operators in the reproducing
kernel Hilbert spaces. Using these empirical measures we develop forward
stepwise (high-dimensional) nonlinear variable selection algorithms. We show
that our algorithm, using the graph-based estimator, yields a provably
consistent model-free variable selection procedure, even in the
high-dimensional regime when the number of covariates grows exponentially with
the sample size, under suitable sparsity assumptions. Extensive simulation and
real-data examples illustrate the superior performance of our methods compared
to existing procedures. The recent conditional dependence measure proposed by
Azadkia and Chatterjee (2019) can be viewed as a special case of our general
framework.
| stat |
Single-stage, three-arm, adaptive test strategies for non-inferiority
trials with an unstable reference | For indications where only unstable reference treatments are available and
use of placebo is ethically justified, three-arm `gold standard' designs with
an experimental, reference and placebo arm are recommended for non-inferiority
trials. In such designs, the demonstration of efficacy of the reference or
experimental treatment is a requirement. They have the disadvantage that only
little can be concluded from the trial if the reference fails to be
efficacious. To overcome this, we investigate novel single-stage, adaptive test
strategies where non-inferiority is tested only if the reference shows
sufficient efficacy and otherwise $\delta$-superiority of the experimental
treatment over placebo is tested. With a properly chosen superiority margin,
$\delta$-superiority indirectly shows non-inferiority. We optimize the sample
size for several decision rules and find that the natural, data driven test
strategy, which tests non-inferiority if the reference's efficacy test is
significant, leads to the smallest overall and placebo sample sizes. We proof
that under specific constraints on the sample sizes, this procedure controls
the family-wise error rate. All optimal sample sizes are found to meet this
constraint. We finally show how to account for a relevant placebo drop-out rate
in an efficient way and apply the new test strategy to a real life data set.
| stat |
A Nonparametric Method for Value Function Guided Subgroup Identification
via Gradient Tree Boosting for Censored Survival Data | In randomized clinical trials with survival outcome, there has been an
increasing interest in subgroup identification based on baseline genomic,
proteomic markers or clinical characteristics. Some of the existing methods
identify subgroups that benefit substantially from the experimental treatment
by directly modeling outcomes or treatment effect. When the goal is to find an
optimal treatment for a given patient rather than finding the right patient for
a given treatment, methods under the individualized treatment regime framework
estimate an individualized treatment rule that would lead to the best expected
clinical outcome as measured by a value function. Connecting the concept of
value function to subgroup identification, we propose a nonparametric method
that searches for subgroup membership scores by maximizing a value function
that directly reflects the subgroup-treatment interaction effect based on
restricted mean survival time. A gradient tree boosting algorithm is proposed
to search for the individual subgroup membership scores. We conduct simulation
studies to evaluate the performance of the proposed method and an application
to an AIDS clinical trial is performed for illustration.
| stat |
Bellman filtering for state-space models | This article presents a filter for state-space models based on Bellman's
dynamic programming principle applied to the mode estimator. The proposed
Bellman filter (BF) generalises the Kalman filter (KF) including its extended
and iterated versions, while remaining equally inexpensive computationally. The
BF is also (unlike the KF) robust under heavy-tailed observation noise and
applicable to a wider range of (nonlinear and non-Gaussian) models, involving
e.g. count, intensity, duration, volatility and dependence. (Hyper)parameters
are estimated by numerically maximising a BF-implied log-likelihood
decomposition, which is an alternative to the classic prediction-error
decomposition for linear Gaussian models. Simulation studies reveal that the BF
performs on par with (or even outperforms) state-of-the-art importance-sampling
techniques, while requiring a fraction of the computational cost, being
straightforward to implement and offering full scalability to higher
dimensional state spaces.
| stat |
Sub-sampling for Efficient Non-Parametric Bandit Exploration | In this paper we propose the first multi-armed bandit algorithm based on
re-sampling that achieves asymptotically optimal regret simultaneously for
different families of arms (namely Bernoulli, Gaussian and Poisson
distributions). Unlike Thompson Sampling which requires to specify a different
prior to be optimal in each case, our proposal RB-SDA does not need any
distribution-dependent tuning. RB-SDA belongs to the family of Sub-sampling
Duelling Algorithms (SDA) which combines the sub-sampling idea first used by
the BESA [1] and SSMC [2] algorithms with different sub-sampling schemes. In
particular, RB-SDA uses Random Block sampling. We perform an experimental study
assessing the flexibility and robustness of this promising novel approach for
exploration in bandit models.
| stat |
Enforcing stationarity through the prior in vector autoregressions | Stationarity is a very common assumption in time series analysis. A vector
autoregressive (VAR) process is stationary if and only if the roots of its
characteristic equation lie outside the unit circle, constraining the
autoregressive coefficient matrices to lie in the stationary region. However,
the stationary region has a highly complex geometry which impedes specification
of a prior distribution. In this work, an unconstrained reparameterisation of a
stationary VAR model is presented. The new parameters are based on partial
autocorrelation matrices, which are interpretable, and can be transformed
bijectively to the space of unconstrained square matrices. This transformation
preserves various structural forms of the partial autocorrelation matrices and
readily facilitates specification of a prior. Properties of this prior are
described along with an important special case which is exchangeable with
respect to the order of the elements in the observation vector. Posterior
inference and computation are described and implemented using Hamiltonian Monte
Carlo via Stan. The prior and inferential procedures are illustrated with an
application to a macroeconomic time series which highlights the benefits of
enforcing stationarity.
| stat |
Balancing Statistical and Computational Precision and Applications to
Penalized Linear Regression with Group Sparsity | Due to technological advances, large and high-dimensional data have become
the rule rather than the exception. Methods that allow for feature selection
with such data are thus highly sought after, in particular, since standard
methods, such as cross-validated lasso and group-lasso, can be challenging both
computationally and mathematically. In this paper, we propose a novel approach
to feature selection and group feature selection in linear regression. It
consists of simple optimization steps and tests, which makes it computationally
more efficient than standard approaches and suitable even for very large data
sets. Moreover, it satisfies sharp guarantees for estimation and feature
selection in terms of oracle inequalities. We thus expect that our contribution
can help to leverage the increasing volume of data in Biology, Public Health,
Astronomy, Economics, and other fields.
| stat |
Learning with CVaR-based feedback under potentially heavy tails | We study learning algorithms that seek to minimize the conditional
value-at-risk (CVaR), when all the learner knows is that the losses incurred
may be heavy-tailed. We begin by studying a general-purpose estimator of CVaR
for potentially heavy-tailed random variables, which is easy to implement in
practice, and requires nothing more than finite variance and a distribution
function that does not change too fast or slow around just the quantile of
interest. With this estimator in hand, we then derive a new learning algorithm
which robustly chooses among candidates produced by stochastic gradient-driven
sub-processes. For this procedure we provide high-probability excess CVaR
bounds, and to complement the theory we conduct empirical tests of the
underlying CVaR estimator and the learning algorithm derived from it.
| stat |
Prediction based on conditional distributions of vine copulas | Vine copulas are a flexible tool for multivariate non-Gaussian distributions.
For data from an observational study where the explanatory variables and
response variables are measured together, a proposed vine copula regression
method uses regular vines and handles mixed continuous and discrete variables.
This method can efficiently compute the conditional distribution of the
response variable given the explanatory variables. The performance of the
proposed method is evaluated on simulated data sets and a real data set. The
experiments demonstrate that the vine copula regression method is superior to
linear regression in making inferences with conditional heteroscedasticity.
| stat |
Communication-Efficient Distributed Optimization in Networks with
Gradient Tracking and Variance Reduction | There is growing interest in large-scale machine learning and optimization
over decentralized networks, e.g. in the context of multi-agent learning and
federated learning. Due to the imminent need to alleviate the communication
burden, the investigation of communication-efficient distributed optimization
algorithms - particularly for empirical risk minimization - has flourished in
recent years. A large fraction of these algorithms have been developed for the
master/slave setting, relying on a central parameter server that can
communicate with all agents. This paper focuses on distributed optimization
over networks, or decentralized optimization, where each agent is only allowed
to aggregate information from its neighbors. By properly adjusting the global
gradient estimate via local averaging in conjunction with proper correction, we
develop a communication-efficient approximate Newton-type method Network-DANE,
which generalizes DANE to the decentralized scenarios. Our key ideas can be
applied in a systematic manner to obtain decentralized versions of other
master/slave distributed algorithms. A notable development is
Network-SVRG/SARAH, which employs variance reduction to further accelerate
local computation. We establish linear convergence of Network-DANE and
Network-SVRG for strongly convex losses, and Network-SARAH for quadratic
losses, which shed light on the impacts of data homogeneity, network
connectivity, and local averaging upon the rate of convergence. We further
extend Network-DANE to composite optimization by allowing a nonsmooth penalty
term. Numerical evidence is provided to demonstrate the appealing performance
of our algorithms over competitive baselines, in terms of both communication
and computation efficiency. Our work suggests that performing a certain amount
of local communications and computations per iteration can substantially
improve the overall efficiency.
| stat |
The Generalization Error of the Minimum-norm Solutions for
Over-parameterized Neural Networks | We study the generalization properties of minimum-norm solutions for three
over-parametrized machine learning models including the random feature model,
the two-layer neural network model and the residual network model. We proved
that for all three models, the generalization error for the minimum-norm
solution is comparable to the Monte Carlo rate, up to some logarithmic terms,
as long as the models are sufficiently over-parametrized.
| stat |
Generative Modeling by Inclusive Neural Random Fields with Applications
in Image Generation and Anomaly Detection | Neural random fields (NRFs), referring to a class of generative models that
use neural networks to implement potential functions in random fields (a.k.a.
energy-based models), are not new but receive less attention with slow
progress. Different from various directed graphical models such as generative
adversarial networks (GANs), NRFs provide an interesting family of undirected
graphical models for generative modeling. In this paper we propose a new
approach, the inclusive-NRF approach, to learning NRFs for continuous data
(e.g. images), by introducing inclusive-divergence minimized auxiliary
generators and developing stochastic gradient sampling in an augmented space.
Based on the new approach, specific inclusive-NRF models are developed and
thoroughly evaluated in two important generative modeling applications - image
generation and anomaly detection. The proposed models consistently improve over
state-of-the-art results in both applications. Remarkably, in addition to
superior sample generation, one additional benefit of our inclusive-NRF
approach is that, unlike GANs, it can directly provide (unnormalized) density
estimate for sample evaluation. With these contributions and results, this
paper significantly advances the learning and applications of NRFs to a new
level, both theoretically and empirically, which have never been obtained
before.
| stat |
Surrogate Locally-Interpretable Models with Supervised Machine Learning
Algorithms | Supervised Machine Learning (SML) algorithms, such as Gradient Boosting,
Random Forest, and Neural Networks, have become popular in recent years due to
their superior predictive performance over traditional statistical methods.
However, their complexity makes the results hard to interpret without
additional tools. There has been a lot of recent work in developing global and
local diagnostics for interpreting SML models. In this paper, we propose a
locally-interpretable model that takes the fitted ML response surface,
partitions the predictor space using model-based regression trees, and fits
interpretable main-effects models at each of the nodes. We adapt the algorithm
to be efficient in dealing with high-dimensional predictors. While the main
focus is on interpretability, the resulting surrogate model also has reasonably
good predictive performance.
| stat |
On Consistency of Graph-based Semi-supervised Learning | Graph-based semi-supervised learning is one of the most popular methods in
machine learning. Some of its theoretical properties such as bounds for the
generalization error and the convergence of the graph Laplacian regularizer
have been studied in computer science and statistics literatures. However, a
fundamental statistical property, the consistency of the estimator from this
method has not been proved. In this article, we study the consistency problem
under a non-parametric framework. We prove the consistency of graph-based
learning in the case that the estimated scores are enforced to be equal to the
observed responses for the labeled data. The sample sizes of both labeled and
unlabeled data are allowed to grow in this result. When the estimated scores
are not required to be equal to the observed responses, a tuning parameter is
used to balance the loss function and the graph Laplacian regularizer. We give
a counterexample demonstrating that the estimator for this case can be
inconsistent. The theoretical findings are supported by numerical studies.
| stat |
Deep Fundamental Factor Models | Deep fundamental factor models are developed to automatically capture
non-linearity and interaction effects in factor modeling. Uncertainty
quantification provides interpretability with interval estimation, ranking of
factor importances and estimation of interaction effects. With no hidden layers
we recover a linear factor model and for one or more hidden layers, uncertainty
bands for the sensitivity to each input naturally arise from the network
weights. Using 3290 assets in the Russell 1000 index over a period of December
1989 to January 2018, we assess a 49 factor model and generate information
ratios that are approximately 1.5x greater than the OLS factor model.
Furthermore, we compare our deep fundamental factor model with a quadratic
LASSO model and demonstrate the superior performance and robustness to
outliers. The Python source code and the data used for this study are provided.
| stat |
Nonparametric Bayesian Instrumental Variable Analysis: Evaluating
Heterogeneous Effects of Coronary Arterial Access Site Strategies | Percutaneous coronary interventions (PCIs) are nonsurgical procedures to open
blocked blood vessels to the heart, frequently using a catheter to place a
stent. The catheter can be inserted into the blood vessels using an artery in
the groin or an artery in the wrist. Because clinical trials have indicated
that access via the wrist may result in fewer post procedure complications,
shortening the length of stay, and ultimately cost less than groin access,
adoption of access via the wrist has been encouraged. However, patients treated
in usual care are likely to differ from those participating in clinical trials,
and there is reason to believe that the effectiveness of wrist access may
differ between males and females. Moreover, the choice of artery access
strategy is likely to be influenced by patient or physician unmeasured factors.
To study the effectiveness of the two artery access site strategies on
hospitalization charges, we use data from a state-mandated clinical registry
including 7,963 patients undergoing PCI. A hierarchical Bayesian
likelihood-based instrumental variable analysis under a latent index modeling
framework is introduced to jointly model outcomes and treatment status. Our
approach accounts for unobserved heterogeneity via a latent factor structure,
and permits nonparametric error distributions with Dirichlet process mixture
models. Our results demonstrate that artery access in the wrist reduces
hospitalization charges compared to access in the groin, with higher mean
reduction for male patients.
| stat |
Extremal Dependence Concepts | The probabilistic characterization of the relationship between two or more
random variables calls for a notion of dependence. Dependence modeling leads to
mathematical and statistical challenges, and recent developments in extremal
dependence concepts have drawn a lot of attention to probability and its
applications in several disciplines. The aim of this paper is to review various
concepts of extremal positive and negative dependence, including several
recently established results, reconstruct their history, link them to
probabilistic optimization problems, and provide a list of open questions in
this area. While the concept of extremal positive dependence is agreed upon for
random vectors of arbitrary dimensions, various notions of extremal negative
dependence arise when more than two random variables are involved. We review
existing popular concepts of extremal negative dependence given in literature
and introduce a novel notion, which in a general sense includes the existing
ones as particular cases. Even if much of the literature on dependence is
focused on positive dependence, we show that negative dependence plays an
equally important role in the solution of many optimization problems. While the
most popular tool used nowadays to model dependence is that of a copula
function, in this paper we use the equivalent concept of a set of
rearrangements. This is not only for historical reasons. Rearrangement
functions describe the relationship between random variables in a completely
deterministic way, allow a deeper understanding of dependence itself, and have
several advantages on the approximation of solutions in a broad class of
optimization problems.
| stat |
Learning with Good Feature Representations in Bandits and in RL with a
Generative Model | The construction by Du et al. (2019) implies that even if a learner is given
linear features in $\mathbb R^d$ that approximate the rewards in a bandit with
a uniform error of $\epsilon$, then searching for an action that is optimal up
to $O(\epsilon)$ requires examining essentially all actions. We use the
Kiefer-Wolfowitz theorem to prove a positive result that by checking only a few
actions, a learner can always find an action that is suboptimal with an error
of at most $O(\epsilon \sqrt{d})$. Thus, features are useful when the
approximation error is small relative to the dimensionality of the features.
The idea is applied to stochastic bandits and reinforcement learning with a
generative model where the learner has access to $d$-dimensional linear
features that approximate the action-value functions for all policies to an
accuracy of $\epsilon$. For linear bandits, we prove a bound on the regret of
order $\sqrt{dn \log(k)} + \epsilon n \sqrt{d} \log(n)$ with $k$ the number of
actions and $n$ the horizon. For RL we show that approximate policy iteration
can learn a policy that is optimal up to an additive error of order $\epsilon
\sqrt{d}/(1 - \gamma)^2$ and using $d/(\epsilon^2(1 - \gamma)^4)$ samples from
a generative model. These bounds are independent of the finer details of the
features. We also investigate how the structure of the feature set impacts the
tradeoff between sample complexity and estimation error.
| stat |
On a Bivariate Copula for Modeling Negative Dependence | A new bivariate copula is proposed for modeling negative dependence between
two random variables. We show that it complies with most of the popular notions
of negative dependence reported in the literature and study some of its basic
properties. Specifically, the Spearman's rho and the Kendall's tau for the
proposed copula have a simple one-parameter form with negative values in the
full range. Some important ordering properties comparing the strength of
negative dependence with respect to the parameter involved are considered.
Simple examples of the corresponding bivariate distributions with popular
marginals are presented. Application of the proposed copula is illustrated
using a real data set.
| stat |
Learning with Sets in Multiple Instance Regression Applied to Remote
Sensing | In this paper, we propose a novel approach to tackle the multiple instance
regression (MIR) problem. This problem arises when the data is a collection of
bags, where each bag is made of multiple instances corresponding to the same
unique real-valued label. Our goal is to train a regression model which maps
the instances of an unseen bag to its unique label. This MIR setting is common
to remote sensing applications where there is high variability in the
measurements and low geographical variability in the quantity being estimated.
Our approach, in contrast to most competing methods, does not make the
assumption that there exists a prime instance responsible for the label in each
bag. Instead, we treat each bag as a set (i.e, an unordered sequence) of
instances and learn to map each bag to its unique label by using all the
instances in each bag. This is done by implementing an order-invariant
operation characterized by a particular type of attention mechanism. This
method is very flexible as it does not require domain knowledge nor does it
make any assumptions about the distribution of the instances within each bag.
We test our algorithm on five real world datasets and outperform previous
state-of-the-art on three of the datasets. In addition, we augment our feature
space by adding the moments of each feature for each bag, as extra features,
and show that while the first moments lead to higher accuracy, there is a
diminishing return.
| stat |
A simple study of the correlation effects in the superposition of waves
of electric fields: the emergence of extreme events | In this paper, we study the effects of correlated random phases in the
intensity of a superposition of $N$ wave-fields. Our results suggest that
regardless of whether the phase distribution is continuous or discrete if the
phases are random correlated variables, we must observe a heavier tail
distribution and the emergence of extreme events as the correlation between
phases increases. We believe that such a simple method can be easily applied in
other situations to show the existence of extreme statistical events in the
context of nonlinear complex systems.
| stat |
Bayesian causal inference in probit graphical models | We consider a binary response which is potentially affected by a set of
continuous variables. Of special interest is the causal effect on the response
due to an intervention on a specific variable. The latter can be meaningfully
determined on the basis of observational data through suitable assumptions on
the data generating mechanism. In particular we assume that the joint
distribution obeys the conditional independencies (Markov properties) inherent
in a Directed Acyclic Graph (DAG), and the DAG is given a causal interpretation
through the notion of interventional distribution. We propose a DAG-probit
model where the response is generated by discretization through a random
threshold of a continuous latent variable and the latter, jointly with the
remaining continuous variables, has a distribution belonging to a zero-mean
Gaussian model whose covariance matrix is constrained to satisfy the Markov
properties of the DAG. Our model leads to a natural definition of causal effect
conditionally on a given DAG. Since the DAG which generates the observations is
unknown, we present an efficient MCMC algorithm whose target is the posterior
distribution on the space of DAGs, the Cholesky parameters of the concentration
matrix, and the threshold linking the response to the latent. Our end result is
a Bayesian Model Averaging estimate of the causal effect which incorporates
parameter, as well as model, uncertainty. The methodology is assessed using
simulation experiments and applied to a gene expression data set originating
from breast cancer stem cells.
| stat |
Dimension-robust Function Space MCMC With Neural Network Priors | This paper introduces a new prior on functions spaces which scales more
favourably in the dimension of the function's domain compared to the usual
Karhunen-Lo\'eve function space prior, a property we refer to as
dimension-robustness. The proposed prior is a Bayesian neural network prior,
where each weight and bias has an independent Gaussian prior, but with the key
difference that the variances decrease in the width of the network, such that
the variances form a summable sequence and the infinite width limit neural
network is well defined. We show that our resulting posterior of the unknown
function is amenable to sampling using Hilbert space Markov chain Monte Carlo
methods. These sampling methods are favoured because they are stable under
mesh-refinement, in the sense that the acceptance probability does not shrink
to 0 as more parameters are introduced to better approximate the well-defined
infinite limit. We show that our priors are competitive and have distinct
advantages over other function space priors. Upon defining a suitable
likelihood for continuous value functions in a Bayesian approach to
reinforcement learning, our new prior is used in numerical examples to
illustrate its performance and dimension-robustness.
| stat |
Reinforcement learning and Bayesian data assimilation for model-informed
precision dosing in oncology | Model-informed precision dosing (MIPD) using therapeutic drug/biomarker
monitoring offers the opportunity to significantly improve the efficacy and
safety of drug therapies. Current strategies comprise model-informed dosing
tables or are based on maximum a-posteriori estimates. These approaches,
however, lack a quantification of uncertainty and/or consider only part of the
available patient-specific information. We propose three novel approaches for
MIPD employing Bayesian data assimilation (DA) and/or reinforcement learning
(RL) to control neutropenia, the major dose-limiting side effect in anticancer
chemotherapy. These approaches have the potential to substantially reduce the
incidence of life-threatening grade 4 and subtherapeutic grade 0 neutropenia
compared to existing approaches. We further show that RL allows to gain further
insights by identifying patient factors that drive dose decisions. Due to its
flexibility, the proposed combined DA-RL approach can easily be extended to
integrate multiple endpoints or patient-reported outcomes, thereby promising
important benefits for future personalized therapies.
| stat |
How do mobility restrictions and social distancing during COVID-19
affect the crude oil price? | We develop an air mobility index and use the newly developed Apple's driving
trend index to evaluate the impact of COVID-19 on the crude oil price. We use
quantile regression and stationary and non-stationary extreme value models to
study the impact. We find that both the \textit{air mobility index} and
\textit{driving trend index} significantly influence lower and upper quantiles
as well as the median of the WTI crude oil price. The extreme value model
suggests that an event like COVID-19 may push oil prices to a negative
territory again as the air mobility decreases drastically during such
pandemics.
| stat |
Sequential Adversarial Anomaly Detection for One-Class Event Data | We consider the sequential anomaly detection problem in the one-class setting
when only the anomalous sequences are available and propose an adversarial
sequential detector by solving a minimax problem to find an optimal detector
against the worst-case sequences from a generator. The generator captures the
dependence in sequential events using the marked point process model. The
detector sequentially evaluates the likelihood of a test sequence and compares
it with a time-varying threshold, also learned from data through the minimax
problem. We demonstrate our proposed method's good performance using numerical
experiments on simulations and proprietary large-scale credit card fraud
datasets. The proposed method can generally apply to detecting anomalous
sequences.
| stat |
Ranking and synchronization from pairwise measurements via SVD | Given a measurement graph $G= (V,E)$ and an unknown signal $r \in
\mathbb{R}^n$, we investigate algorithms for recovering $r$ from pairwise
measurements of the form $r_i - r_j$; $\{i,j\} \in E$. This problem arises in a
variety of applications, such as ranking teams in sports data and time
synchronization of distributed networks. Framed in the context of ranking, the
task is to recover the ranking of $n$ teams (induced by $r$) given a small
subset of noisy pairwise rank offsets. We propose a simple SVD-based
algorithmic pipeline for both the problem of time synchronization and ranking.
We provide a detailed theoretical analysis in terms of robustness against both
sampling sparsity and noise perturbations with outliers, using results from
matrix perturbation and random matrix theory. Our theoretical findings are
complemented by a detailed set of numerical experiments on both synthetic and
real data, showcasing the competitiveness of our proposed algorithms with other
state-of-the-art methods.
| stat |
From industry-wide parameters to aircraft-centric on-flight inference:
improving aeronautics performance prediction with machine learning | Aircraft performance models play a key role in airline operations, especially
in planning a fuel-efficient flight. In practice, manufacturers provide
guidelines which are slightly modified throughout the aircraft life cycle via
the tuning of a single factor, enabling better fuel predictions. However this
has limitations, in particular they do not reflect the evolution of each
feature impacting the aircraft performance. Our goal here is to overcome this
limitation. The key contribution of the present article is to foster the use of
machine learning to leverage the massive amounts of data continuously recorded
during flights performed by an aircraft and provide models reflecting its
actual and individual performance. We illustrate our approach by focusing on
the estimation of the drag and lift coefficients from recorded flight data. As
these coefficients are not directly recorded, we resort to aerodynamics
approximations. As a safety check, we provide bounds to assess the accuracy of
both the aerodynamics approximation and the statistical performance of our
approach. We provide numerical results on a collection of machine learning
algorithms. We report excellent accuracy on real-life data and exhibit
empirical evidence to support our modelling, in coherence with aerodynamics
principles.
| stat |
Mape_Maker: A Scenario Creator | We describe algorithms for creating probabilistic scenarios for the situation
when the underlying forecast methodology is modeled as being more (or less)
accurate than it has been historically. Such scenarios can be used in studies
that extend into the future and may need to consider the possibility that
forecast technology will improve. Our approach can also be used to generate
alternative realizations of renewable energy production that are consistent
with historical forecast accuracy, in effect serving as a method for creating
families of realistic alternatives -- which are often critical in
simulation-based analysis methodologies
| stat |
A table of short-period Tausworthe generators for Markov chain
quasi-Monte Carlo | We consider the problem of estimating expectations by using Markov chain
Monte Carlo methods and improving the accuracy by replacing IID uniform random
points with quasi-Monte Carlo (QMC) points. Recently, it has been shown that
Markov chain QMC remains consistent when the driving sequences are completely
uniformly distributed (CUD). However, the definition of CUD sequences is not
constructive, so an implementation method using short-period Tausworthe
generators (i.e., linear feedback shift register generators over the
two-element field) that approximate CUD sequences has been proposed. In this
paper, we conduct an exhaustive search of short-period Tausworthe generators
for Markov chain QMC in terms of the $t$-value, which is a criterion of
uniformity widely used in the study of QMC methods. We provide a parameter
table of Tausworthe generators and show the effectiveness in numerical examples
using Gibbs sampling.
| stat |
Maximum Correntropy Criterion with Variable Center | Correntropy is a local similarity measure defined in kernel space and the
maximum correntropy criterion (MCC) has been successfully applied in many areas
of signal processing and machine learning in recent years. The kernel function
in correntropy is usually restricted to the Gaussian function with center
located at zero. However, zero-mean Gaussian function may not be a good choice
for many practical applications. In this study, we propose an extended version
of correntropy, whose center can locate at any position. Accordingly, we
propose a new optimization criterion called maximum correntropy criterion with
variable center (MCC-VC). We also propose an efficient approach to optimize the
kernel width and center location in MCC-VC. Simulation results of regression
with linear in parameters (LIP) models confirm the desirable performance of the
new method.
| stat |
Improving the Hosmer-Lemeshow Goodness-of-Fit Test in Large Models with
Replicated Trials | The Hosmer-Lemeshow (HL) test is a commonly used global goodness-of-fit (GOF)
test that assesses the quality of the overall fit of a logistic regression
model. In this paper, we give results from simulations showing that the type 1
error rate (and hence power) of the HL test decreases as model complexity
grows, provided that the sample size remains fixed and binary replicates are
present in the data. We demonstrate that the generalized version of the HL test
by Surjanovic et al. (2020) can offer some protection against this power loss.
We conclude with a brief discussion explaining the behaviour of the HL test,
along with some guidance on how to choose between the two tests.
| stat |
Imputation estimators for unnormalized models with missing data | Several statistical models are given in the form of unnormalized densities,
and calculation of the normalization constant is intractable. We propose
estimation methods for such unnormalized models with missing data. The key
concept is to combine imputation techniques with estimators for unnormalized
models including noise contrastive estimation and score matching. In addition,
we derive asymptotic distributions of the proposed estimators and construct
confidence intervals. Simulation results with truncated Gaussian graphical
models and the application to real data of wind direction reveal that the
proposed methods effectively enable statistical inference with unnormalized
models from missing data.
| stat |
Predicting with Proxies: Transfer Learning in High Dimension | Predictive analytics is increasingly used to guide decision-making in many
applications. However, in practice, we often have limited data on the true
predictive task of interest, and must instead rely on more abundant data on a
closely-related proxy predictive task. For example, e-commerce platforms use
abundant customer click data (proxy) to make product recommendations rather
than the relatively sparse customer purchase data (true outcome of interest);
alternatively, hospitals often rely on medical risk scores trained on a
different patient population (proxy) rather than their own patient population
(true cohort of interest) to assign interventions. Yet, not accounting for the
bias in the proxy can lead to sub-optimal decisions. Using real datasets, we
find that this bias can often be captured by a sparse function of the features.
Thus, we propose a novel two-step estimator that uses techniques from
high-dimensional statistics to efficiently combine a large amount of proxy data
and a small amount of true data. We prove upper bounds on the error of our
proposed estimator and lower bounds on several heuristics used by data
scientists; in particular, our proposed estimator can achieve the same accuracy
with exponentially less true data (in the number of features). Our proof relies
on a new LASSO tail inequality for approximately sparse vectors. Finally, we
demonstrate the effectiveness of our approach on e-commerce and healthcare
datasets; in both cases, we achieve significantly better predictive accuracy as
well as managerial insights into the nature of the bias in the proxy data.
| stat |
Causal Inference for Multiple Non-Randomized Treatments using Fractional
Factorial Designs | We explore a framework for addressing causal questions in an observational
setting with multiple treatments. This setting involves attempting to
approximate an experiment from observational data. With multiple treatments,
this experiment would be a factorial design. However, certain treatment
combinations may be so rare that there are no measured outcomes in the observed
data corresponding to them. We propose to conceptualize a hypothetical
fractional factorial experiment instead of a full factorial experiment and lay
out a framework for analysis in this setting. We connect our design-based
methods to standard regression methods. We finish by illustrating our approach
using biomedical data from the 2003-2004 cycle of the National Health and
Nutrition Examination Survey to estimate the effects of four common pesticides
on body mass index.
| stat |
Block Stability for MAP Inference | To understand the empirical success of approximate MAP inference, recent work
(Lang et al., 2018) has shown that some popular approximation algorithms
perform very well when the input instance is stable. The simplest stability
condition assumes that the MAP solution does not change at all when some of the
pairwise potentials are (adversarially) perturbed. Unfortunately, this strong
condition does not seem to be satisfied in practice. In this paper, we
introduce a significantly more relaxed condition that only requires blocks
(portions) of an input instance to be stable. Under this block stability
condition, we prove that the pairwise LP relaxation is persistent on the stable
blocks. We complement our theoretical results with an empirical evaluation of
real-world MAP inference instances from computer vision. We design an algorithm
to find stable blocks, and find that these real instances have large stable
regions. Our work gives a theoretical explanation for the widespread empirical
phenomenon of persistency for this LP relaxation.
| stat |
Discriminator optimal transport | Within a broad class of generative adversarial networks, we show that
discriminator optimization process increases a lower bound of the dual cost
function for the Wasserstein distance between the target distribution $p$ and
the generator distribution $p_G$. It implies that the trained discriminator can
approximate optimal transport (OT) from $p_G$ to $p$.Based on some experiments
and a bit of OT theory, we propose a discriminator optimal transport (DOT)
scheme to improve generated images. We show that it improves inception score
and FID calculated by un-conditional GAN trained by CIFAR-10, STL-10 and a
public pre-trained model of conditional GAN by ImageNet.
| stat |
Fast Inference Procedures for Semivarying Coefficient Models via Local
Averaging | The semivarying coefficient models are widely used in the application of
finance, economics, medical science and many other areas. The functional
coefficients are commonly estimated by local smoothing methods, e.g. local
linear estimator. This implies that one should implement the estimation
procedure for hundreds of times to obtain an estimate of one function. So the
computation cost is very severe. In this paper, we give an insight to the
trade-off between statistical efficiency and computation simplicity, and
proposes a fast inference procedure for semivarying coefficient model. In our
method, the coefficient functions are approximated by piecewise constants,
which is a simple and rough approximation. This makes our estimators easy to
implement and avoid repeat estimation. In this work, we shall show that though
these estimators are not asymptotically optimal, they are efficient enough for
building further inference procedure. Furthermore, three tests are brought out
to check whether certain coefficient is constant. Our results clearly show that
when the room for improving the asymptotic efficiency is limited, a proper
trade-off between statistical efficiency and computation simplicity can be
taken into consideration to improve the performance of the inference procedure.
| stat |
Random weighting in LASSO regression | We establish statistical properties of random-weighting methods in LASSO
regression under different regularization parameters $\lambda_n$ and suitable
regularity conditions. The random-weighting methods in view concern repeated
optimization of a randomized objective function, motivated by the need for
computational approximations to Bayesian posterior sampling. In the context of
LASSO regression, we repeatedly assign analyst-drawn random weights to terms in
the objective function (including the penalty terms), and optimize to obtain a
sample of random-weighting estimators. We show that existing approaches have
conditional model selection consistency and conditional asymptotic normality at
different growth rates of $\lambda_n$ as $n \to \infty$. We propose an
extension to the available random-weighting methods and establish that the
resulting samples attain conditional sparse normality and conditional
consistency in a growing-dimension setting. We find that random-weighting has
both approximate-Bayesian and sampling-theory interpretations. Finally, we
illustrate the proposed methodology via extensive simulation studies and a
benchmark data example.
| stat |
Should Observations be Grouped for Effective Monitoring of Multivariate
Process Variability? | A multivariate dispersion control chart monitors changes in the process
variability of multiple correlated quality characteristics. In this article, we
investigate and compare the performance of charts designed to monitor
variability based on individual and grouped multivariate observations. We
compare one of the most well-known methods for monitoring individual
observations -- a multivariate EWMA chart proposed by Huwang et al -- to
various charts based on grouped observations. In addition, we compare charts
based on monitoring with overlapping and nonoverlapping subgroups. We recommend
using charts based on overlapping subgroups when monitoring with subgroup data.
The effect of subgroup size is also investigated. Steady-state average time to
signal is used as performance measure. We show that monitoring methods based on
individual observations are the quickest in detecting sustained shifts in the
process variability. We use a simulation study to obtain our results and
illustrated these with a case study.
| stat |
A Probabilistic Model for Predicting Shot Success in Football | Football forecasting models traditionally rate teams on past match results,
that is based on the number of goals scored. Goals, however, involve a high
element of chance and thus past results often do not reflect the performances
of the teams. In recent years, it has become increasingly clear that accounting
for other match events such as shots at goal can provide a better indication of
the relative strengths of two teams than the number of goals scored. Forecast
models based on this information have been shown to be successful in
outperforming those based purely on match results. A notable weakness, however,
is that this approach does not take into account differences in the probability
of shot success among teams. A team that is more likely to score from a shot
will need fewer shots to win a match, on average. In this paper, we propose a
simple parametric model to predict the probability of a team scoring, given it
has taken a shot at goal. We show that the resulting forecasts are able to
outperform a model assuming an equal probability of shot success among all
teams. We then show that the model can be combined with predictions of the
number of shots achieved by each team, and can increase the skill of forecasts
of both the match outcome and of whether the total number of goals in a match
will exceed 2.5. We assess the performance of the forecasts alongside two
betting strategies and find mixed evidence for improved performance.
| stat |
Trimmed Constrained Mixed Effects Models: Formulations and Algorithms | Mixed effects (ME) models inform a vast array of problems in the physical and
social sciences, and are pervasive in meta-analysis. We consider ME models
where the random effects component is linear. We then develop an efficient
approach for a broad problem class that allows nonlinear measurements, priors,
and constraints, and finds robust estimates in all of these cases using
trimming in the associated marginal likelihood.
The software accompanying this paper is disseminated as an open-source Python
package called LimeTr. LimeTr is able to recover results more accurately in the
presence of outliers compared to available packages for both standard
longitudinal analysis and meta-analysis, and is also more computationally
efficient than competing robust alternatives. Supplementary materials that
reproduce the simulations, as well as run LimeTr and third party code are
available online. We also present analyses of global health data, where we use
advanced functionality of LimeTr, including constraints to impose monotonicity
and concavity for dose-response relationships. Nonlinear observation models
allow new analyses in place of classic approximations, such as log-linear
models. Robust extensions in all analyses ensure that spurious data points do
not drive our understanding of either mean relationships or between-study
heterogeneity.
| stat |
An Optimal Test for the Additive Model with Discrete or Categorical
Predictors | In multivariate nonparametric regression the additive models are very useful
when a suitable parametric model is difficult to find. The backfitting
algorithm is a powerful tool to estimate the additive components. However, due
to complexity of the estimators, the asymptotic $p$-value of the associated
test is difficult to calculate without a Monte Carlo simulation. Moreover, the
conventional tests assume that the predictor variables are strictly continuous.
In this paper, a new test is introduced for the additive components with
discrete or categorical predictors, where the model may contain continuous
covariates. This method is also applied to the semiparametric regression to
test the goodness-of-fit of the model. These tests are asymptotically optimal
in terms of the rate of convergence, as they can detect a specific class of
contiguous alternatives at a rate of $n^{-1/2}$. An extensive simulation study
is presented to support the theoretical results derived in this paper. Finally,
the method is applied to a real data to model the diamond price based on its
quality attributes and physical measurements.
| stat |
Linear mixed models with endogenous covariates: modeling sequential
treatment effects with application to a mobile health study | Mobile health is a rapidly developing field in which behavioral treatments
are delivered to individuals via wearables or smartphones to facilitate
health-related behavior change. Micro-randomized trials (MRT) are an
experimental design for developing mobile health interventions. In an MRT the
treatments are randomized numerous times for each individual over course of the
trial. Along with assessing treatment effects, behavioral scientists aim to
understand between-person heterogeneity in the treatment effect. A natural
approach is the familiar linear mixed model. However, directly applying linear
mixed models is problematic because potential moderators of the treatment
effect are frequently endogenous---that is, may depend on prior treatment. We
discuss model interpretation and biases that arise in the absence of additional
assumptions when endogenous covariates are included in a linear mixed model. In
particular, when there are endogenous covariates, the coefficients no longer
have the customary marginal interpretation. However, these coefficients still
have a conditional-on-the-random-effect interpretation. We provide an
additional assumption that, if true, allows scientists to use standard software
to fit linear mixed model with endogenous covariates, and person-specific
predictions of effects can be provided. As an illustration, we assess the
effect of activity suggestion in the HeartSteps MRT and analyze the
between-person treatment effect heterogeneity.
| stat |
Non-exchangeable feature allocation models with sublinear growth of the
feature sizes | Feature allocation models are popular models used in different applications
such as unsupervised learning or network modeling. In particular, the Indian
buffet process is a flexible and simple one-parameter feature allocation model
where the number of features grows unboundedly with the number of objects. The
Indian buffet process, like most feature allocation models, satisfies a
symmetry property of exchangeability: the distribution is invariant under
permutation of the objects. While this property is desirable in some cases, it
has some strong implications. Importantly, the number of objects sharing a
particular feature grows linearly with the number of objects. In this article,
we describe a class of non-exchangeable feature allocation models where the
number of objects sharing a given feature grows sublinearly, where the rate can
be controlled by a tuning parameter. We derive the asymptotic properties of the
model, and show that such model provides a better fit and better predictive
performances on various datasets.
| stat |
Survival Forests under Test: Impact of the Proportional Hazards
Assumption on Prognostic and Predictive Forests for ALS Survival | We investigate the effect of the proportional hazards assumption on
prognostic and predictive models of the survival time of patients suffering
from amyotrophic lateral sclerosis (ALS). We theoretically compare the
underlying model formulations of several variants of survival forests and
implementations thereof, including random forests for survival, conditional
inference forests, Ranger, and survival forests with $L_1$ splitting, with two
novel variants, namely distributional and transformation survival forests.
Theoretical considerations explain the low power of log-rank-based splitting in
detecting patterns in non-proportional hazards situations in survival trees and
corresponding forests. This limitation can potentially be overcome by the
alternative split procedures suggested herein. We empirically investigated this
effect using simulation experiments and a re-analysis of the PRO-ACT database
of ALS survival, giving special emphasis to both prognostic and predictive
models.
| stat |
Generalization error bounds for kernel matrix completion and
extrapolation | Prior information can be incorporated in matrix completion to improve
estimation accuracy and extrapolate the missing entries. Reproducing kernel
Hilbert spaces provide tools to leverage the said prior information, and derive
more reliable algorithms. This paper analyzes the generalization error of such
approaches, and presents numerical tests confirming the theoretical results.
| stat |
Local False Discovery Rate Based Methods for Multiple Testing of One-Way
Classified Hypotheses | This paper continues the line of research initiated in
\cite{Liu:Sarkar:Zhao:2016} on developing a novel framework for multiple
testing of hypotheses grouped in a one-way classified form using
hypothesis-specific local false discovery rates (Lfdr's). It is built on an
extension of the standard two-class mixture model from single to multiple
groups, defining hypothesis-specific Lfdr as a function of the conditional Lfdr
for the hypothesis given that it is within a significant group and the Lfdr for
the group itself and involving a new parameter that measures grouping effect.
This definition captures the underlying group structure for the hypotheses
belonging to a group more effectively than the standard two-class mixture
model. Two new Lfdr based methods, possessing meaningful optimalities, are
produced in their oracle forms. One, designed to control false discoveries
across the entire collection of hypotheses, is proposed as a powerful
alternative to simply pooling all the hypotheses into a single group and using
commonly used Lfdr based method under the standard single-group two-class
mixture model. The other is proposed as an Lfdr analog of the method of
\cite{Benjamini:Bogomolov:2014} for selective inference. It controls Lfdr based
measure of false discoveries associated with selecting groups concurrently with
controlling the average of within-group false discovery proportions across the
selected groups. Simulation studies and real-data application show that our
proposed methods are often more powerful than their relevant competitors.
| stat |
Contextual Constrained Learning for Dose-Finding Clinical Trials | Clinical trials in the medical domain are constrained by budgets. The number
of patients that can be recruited is therefore limited. When a patient
population is heterogeneous, this creates difficulties in learning subgroup
specific responses to a particular drug and especially for a variety of
dosages. In addition, patient recruitment can be difficult by the fact that
clinical trials do not aim to provide a benefit to any given patient in the
trial. In this paper, we propose C3T-Budget, a contextual constrained clinical
trial algorithm for dose-finding under both budget and safety constraints. The
algorithm aims to maximize drug efficacy within the clinical trial while also
learning about the drug being tested. C3T-Budget recruits patients with
consideration of the remaining budget, the remaining time, and the
characteristics of each group, such as the population distribution, estimated
expected efficacy, and estimation credibility. In addition, the algorithm aims
to avoid unsafe dosages. These characteristics are further illustrated in a
simulated clinical trial study, which corroborates the theoretical analysis and
demonstrates an efficient budget usage as well as a balanced learning-treatment
trade-off.
| stat |
Low-rank Tensor Bandits | In recent years, multi-dimensional online decision making has been playing a
crucial role in many practical applications such as online recommendation and
digital marketing. To solve it, we introduce stochastic low-rank tensor
bandits, a class of bandits whose mean rewards can be represented as a low-rank
tensor. We propose two learning algorithms, tensor epoch-greedy and tensor
elimination, and develop finite-time regret bounds for them. We observe that
tensor elimination has an optimal dependency on the time horizon, while tensor
epoch-greedy has a sharper dependency on tensor dimensions. Numerical
experiments further back up these theoretical findings and show that our
algorithms outperform various state-of-the-art approaches that ignore the
tensor low-rank structure.
| stat |
Robust Generative Restricted Kernel Machines using Weighted Conjugate
Feature Duality | Interest in generative models has grown tremendously in the past decade.
However, their training performance can be adversely affected by contamination,
where outliers are encoded in the representation of the model. This results in
the generation of noisy data. In this paper, we introduce weighted conjugate
feature duality in the framework of Restricted Kernel Machines (RKMs). The RKM
formulation allows for an easy integration of methods from classical robust
statistics. This formulation is used to fine-tune the latent space of
generative RKMs using a weighting function based on the Minimum Covariance
Determinant, which is a highly robust estimator of multivariate location and
scatter. Experiments show that the weighted RKM is capable of generating clean
images when contamination is present in the training data. We further show that
the robust method also preserves uncorrelated feature learning through
qualitative and quantitative experiments on standard datasets.
| stat |
Adaptive Dimension Reduction to Accelerate Infinite-Dimensional
Geometric Markov Chain Monte Carlo | Bayesian inverse problems highly rely on efficient and effective inference
methods for uncertainty quantification (UQ). Infinite-dimensional MCMC
algorithms, directly defined on function spaces, are robust under refinement of
physical models. Recent development of this class of algorithms has started to
incorporate the geometry of the posterior informed by data so that they are
capable of exploring complex probability structures. However, the required
geometric quantities are usually expensive to obtain in high dimensions. On the
other hand, most geometric information of the unknown parameter space in this
setting is concentrated in an \emph{intrinsic} finite-dimensional subspace. To
mitigate the computational intensity and scale up the applications of
infinite-dimensional geometric MCMC ($\infty$-GMC), we apply geometry-informed
algorithms to the intrinsic subspace to probe its complex structure, and
simpler methods like preconditioned Crank-Nicolson (pCN) to its geometry-flat
complementary subspace. In this work, we take advantage of dimension reduction
techniques to accelerate the original $\infty$-GMC algorithms. More
specifically, partial spectral decomposition of the (prior or posterior)
covariance operator is used to identify certain number of principal
eigen-directions as a basis for the intrinsic subspace. The combination of
dimension-independent algorithms, geometric information, and dimension
reduction yields more efficient implementation, \emph{(adaptive)
dimension-reduced infinite-dimensional geometric MCMC}. With a small amount of
computational overhead, we can achieve over 70 times speed-up compared to pCN
using a simulated elliptic inverse problem and an inverse problem involving
turbulent combustion. A number of error bounds comparing various MCMC proposals
are presented to predict the asymptotic behavior of the proposed
dimension-reduced algorithms.
| stat |
A Regularized Framework for Sparse and Structured Neural Attention | Modern neural networks are often augmented with an attention mechanism, which
tells the network where to focus within the input. We propose in this paper a
new framework for sparse and structured attention, building upon a smoothed max
operator. We show that the gradient of this operator defines a mapping from
real values to probabilities, suitable as an attention mechanism. Our framework
includes softmax and a slight generalization of the recently-proposed sparsemax
as special cases. However, we also show how our framework can incorporate
modern structured penalties, resulting in more interpretable attention
mechanisms, that focus on entire segments or groups of an input. We derive
efficient algorithms to compute the forward and backward passes of our
attention mechanisms, enabling their use in a neural network trained with
backpropagation. To showcase their potential as a drop-in replacement for
existing ones, we evaluate our attention mechanisms on three large-scale tasks:
textual entailment, machine translation, and sentence summarization. Our
attention mechanisms improve interpretability without sacrificing performance;
notably, on textual entailment and summarization, we outperform the standard
attention mechanisms based on softmax and sparsemax.
| stat |
Insight into bias in time-stratified case-crossover studies | The use of case-crossover designs has become widespread in epidemiological
and medical investigations of transient associations. However, the most popular
reference-select strategy, the time-stratified schema, is not a suitable
solution for controlling bias in case-crossover studies. To prove this, we
conducted a time series decomposition for daily ozone (O3) records; scrutinized
the ability of the time-stratified schema on controlling the yearly, monthly
and weekly time trends; and found it failed on controlling the weekly time
trend. Based on this finding, we proposed a new logistic regression approach in
which we did adjustment for the weekly time trend. A comparison between the
traditional model and the proposed method was done by simulation. An empirical
study was conducted to explore potential associations between air pollutants
and AMI hospitalizations. In summary, time-stratified schema provide effective
control on yearly and monthly time trends but not on weekly time trend.
Therefore, the estimation from the traditional logistical regression basically
reveals the effect of weekly time trend, instead of the transient effect. In
contrast, the proposed logistic regression with adjustment for weekly time
trend can effectively eliminate system bias in case-crossover studies.
| stat |
A Generalized Correlated Random Walk Converging to Fractional Brownian
Motion | We propose a new algorithm to generate a fractional Brownian motion, with a
given Hurst parameter, 1/2<H<1 using the correlated Bernoulli random variables
with parameter p; having a certain density. This density is constructed using
the link between the correlation of multivariate Gaussian random variables and
the correlation of their dichotomized binary variables and the relation between
the correlation coefficient and the persistence parameter. We prove that the
normalized sum of trajectories of this proposed random walk yields a Gaussian
process whose scaling limit is the desired fractional Brownian motion with the
given Hurst parameter, 1/2<H<1
| stat |
Optimal Designs for Second-Order Interactions in Paired Comparison
Experiments with Binary Attributes | In paired comparison experiments respondents usually evaluate pairs of
competing options. For this situation we introduce an appropriate model and
derive optimal designs in the presence of second-order interactions when all
attributes are dichotomous.
| stat |
Preconditioned training of normalizing flows for variational inference
in inverse problems | Obtaining samples from the posterior distribution of inverse problems with
expensive forward operators is challenging especially when the unknowns involve
the strongly heterogeneous Earth. To meet these challenges, we propose a
preconditioning scheme involving a conditional normalizing flow (NF) capable of
sampling from a low-fidelity posterior distribution directly. This conditional
NF is used to speed up the training of the high-fidelity objective involving
minimization of the Kullback-Leibler divergence between the predicted and the
desired high-fidelity posterior density for indirect measurements at hand. To
minimize costs associated with the forward operator, we initialize the
high-fidelity NF with the weights of the pretrained low-fidelity NF, which is
trained beforehand on available model and data pairs. Our numerical
experiments, including a 2D toy and a seismic compressed sensing example,
demonstrate that thanks to the preconditioning considerable speed-ups are
achievable compared to training NFs from scratch.
| stat |
New stochastic highway capacity estimation method and why product limit
method is unsuitable | Kaplan-Meier estimate, commonly known as product limit method (PLM), and
maximum likelihood estimate (MLE) methods in general are often cited as means
of stochastic highway capacity estimation. This article discusses their
unsuitability for such application as properties of traffic flow do not meet
the assumptions for use of the methods. They assume the observed subject has a
history which it went through and did not fail. However, due to its nature,
each traffic flow measurement behaves as a separate subject which did not go
through all the lower levels of intensity (did not "age"). An alternative
method is proposed. It fits the resulting cumulative frequency of breakdowns
with respect to the traffic flow intensity leading to the breakdown instead of
directly estimating the underlying probability distribution of capacity.
Analyses of accuracy and sensitivity to data quantity and censoring rate of the
new method are provided along with comparison to the PLM. The results prove
unsuitability of the PLM and MLE methods in general. The new method is then
used in a case study which compares capacity of a work-zone with and without a
traffic flow speed harmonisation system installed. The results confirm positive
effect of harmonisation on capacity.
| stat |
Stochastic optimization for numerical evaluation of imprecise
probabilities | In applications of imprecise probability, analysts must compute lower (or
upper) expectations, defined as the infimum of an expectation over a set of
parameter values. Monte Carlo methods consistently approximate expectations at
fixed parameter values, but can be costly to implement in grid search to locate
minima over large subsets of the parameter space. We investigate the use of
stochastic iterative root-finding methods for efficiently computing lower
expectations. In two examples we illustrate the use of various stochastic
approximation methods, and demonstrate their superior performance in comparison
to grid search.
| stat |
A Limited-Capacity Minimax Theorem for Non-Convex Games or: How I
Learned to Stop Worrying about Mixed-Nash and Love Neural Nets | Adversarial training, a special case of multi-objective optimization, is an
increasingly prevalent machine learning technique: some of its most notable
applications include GAN-based generative modeling and self-play techniques in
reinforcement learning which have been applied to complex games such as Go or
Poker. In practice, a \emph{single} pair of networks is typically trained in
order to find an approximate equilibrium of a highly nonconcave-nonconvex
adversarial problem. However, while a classic result in game theory states such
an equilibrium exists in concave-convex games, there is no analogous guarantee
if the payoff is nonconcave-nonconvex. Our main contribution is to provide an
approximate minimax theorem for a large class of games where the players pick
neural networks including WGAN, StarCraft II, and Blotto Game. Our findings
rely on the fact that despite being nonconcave-nonconvex with respect to the
neural networks parameters, these games are concave-convex with respect to the
actual models (e.g., functions or distributions) represented by these neural
networks.
| stat |
Structure retrieval from 4D-STEM: statistical analysis of potential
pitfalls in high-dimensional data | Four-dimensional scanning transmission electron microscopy (4D-STEM) is one
of the most rapidly growing modes of electron microscopy imaging. The advent of
fast pixelated cameras and the associated data infrastructure have greatly
accelerated this process. Yet conversion of the 4D datasets into physically
meaningful structure images in real-space remains an open issue. In this work,
we demonstrate that, it is possible to systematically create filters that will
affect the apparent resolution or even qualitative features of the real-space
structure image, reconstructing artificially generated patterns. As initial
efforts, we explore statistical model selection algorithms, aiming for
robustness and reliability of estimated filters. This statistical model
selection analysis demonstrates the need for regularization and
cross-validation of inversion methods to robustly recover structure from
high-dimensional diffraction datasets.
| stat |
rags2ridges: A One-Stop-Shop for Graphical Modeling of High-Dimensional
Precision Matrices | A graphical model is an undirected network representing the conditional
independence properties between random variables. Graphical modeling has become
part and parcel of systems or network approaches to multivariate data, in
particular when the variable dimension exceeds the observation dimension.
rags2ridges is an R package for graphical modeling of high-dimensional
precision matrices. It provides a modular framework for the extraction,
visualization, and analysis of Gaussian graphical models from high-dimensional
data. Moreover, it can handle the incorporation of prior information as well as
multiple heterogeneous data classes. As such, it provides a one-stop-shop for
graphical modeling of high-dimensional precision matrices. The functionality of
the package is illustrated with an example dataset pertaining to blood-based
metabolite measurements in persons suffering from Alzheimer's Disease.
| stat |
Mobility-on-demand versus fixed-route transit systems: an evaluation of
traveler preferences in low-income communities | Emerging transportation technologies, such as ride-hailing and autonomous
vehicles, are disrupting the transportation sector and transforming public
transit. Some transit observers envision future public transit to be integrated
transit systems with fixed-route services running along major corridors and
on-demand ridesharing services covering lower-density areas. A switch from a
conventional fixed-route service model to this kind of integrated
mobility-on-demand transit system, however, may elicit varied responses from
local residents. This paper evaluates traveler preferences for a proposed
integrated mobility-on-demand transit system versus the existing fixed-route
system, with a particular focus on disadvantaged travelers. We conducted a
survey in two low-resource communities in the United States, namely, Detroit
and Ypsilanti, Michigan. A majority of survey respondents preferred a
mobility-on-demand transit system over a fixed-route one. Based on ordered
logit model outputs, we found a stronger preference for mobility-on-demand
transit among males, college graduates, individuals who have never heard of or
used ride-hailing before, and individuals who currently receive inferior
transit services. By contrast, preferences varied little by age, income, race,
or disability status. The most important benefit of a mobility-on-demand
transit system perceived by the survey respondents is enhanced transit
accessibility to different destinations, whereas their major concerns include
the need to actively request rides, possible transit-fare increases, and
potential technological failures. Addressing the concerns of female riders and
accommodating the needs of less technology-proficient individuals should be
major priorities for transit agencies that are considering mobility-on-demand
initiatives.
| stat |
Covariate adjustment in continuous biomarker assessment | Continuous biomarkers are common for disease screening and diagnosis. To
reach a dichotomous clinical decision, a threshold would be imposed to
distinguish subjects with disease from non-diseased individuals. Among various
performance metrics for a continuous biomarker, specificity at a controlled
sensitivity level (or vice versa) is often desirable for clinical utility since
it directly targets where the clinical test is intended to operate. Covariates,
such as age, race, and sample collection, could impact the controlled
sensitivity level in subpopulations and may also confound the association
between biomarker and disease status. Therefore, covariate adjustment is
important in such biomarker evaluation. In this paper, we suggest to adopt a
parsimonious quantile regression model for the diseased population, locally at
the controlled sensitivity level, and assess specificity with
covariate-specific control of the sensitivity. Variance estimates are obtained
from a sample-based approach and bootstrap. Furthermore, our proposed local
model extends readily to a global one for covariate adjustment for the receiver
operating characteristic (ROC) curve over the sensitivity continuum. We
demonstrate computational efficiency of this proposed method and restore the
inherent monotonicity in the estimated covariate-adjusted ROC curve. The
asymptotic properties of the proposed estimators are established. Simulation
studies show favorable performance of the proposal. Finally, we illustrate our
method in biomarker evaluation for aggressive prostate cancer.
| stat |
RNN-based counterfactual prediction, with an application to homestead
policy and public schooling | This paper proposes a method for estimating the effect of a policy
intervention on an outcome over time. We train recurrent neural networks (RNNs)
on the history of control unit outcomes to learn a useful representation for
predicting future outcomes. The learned representation of control units is then
applied to the treated units for predicting counterfactual outcomes. RNNs are
specifically structured to exploit temporal dependencies in panel data, and are
able to learn negative and nonlinear interactions between control unit
outcomes. We apply the method to the problem of estimating the long-run impact
of U.S. homestead policy on public school spending.
| stat |
Monte Carlo integration of non-differentiable functions on
$[0,1]^\iota$, $\iota=1,\dots,d$, using a single determinantal point pattern
defined on $[0,1]^d$ | This paper concerns the use of a particular class of determinantal point
processes (DPP), a class of repulsive spatial point processes, for Monte Carlo
integration. Let $d\ge 1$, $I\subseteq \overline d=\{1,\dots,d\}$ with
$\iota=|I|$. Using a single set of $N$ quadrature points $\{u_1,\dots,u_N\}$
defined, once for all, in dimension $d$ from the realization of the DPP model,
we investigate "minimal" assumptions on the integrand in order to obtain
unbiased Monte Carlo estimates of $\mu(f_I)=\int_{[0,1]^\iota} f_I(u)
\mathrm{d} u$ for any known $\iota$-dimensional integrable function on
$[0,1]^\iota$. In particular, we show that the resulting estimator has variance
with order $N^{-1-(2s\wedge 1)/d}$ when the integrand belongs to some Sobolev
space with regularity $s > 0$. When $s>1/2$ (which includes a large class of
non-differentiable functions), the variance is asymptotically explicit and the
estimator is shown to satisfy a Central Limit Theorem.
| stat |
New ITEM response models: application to school bullying data | School bullying victimization is a variable that cannot be measured directly.
Taking into account that this variable has a lower bound, given by the absence
of bullying victimization, this paper proposes IRT logistic models, where the
latent parameter ranges from $0$ to $\infty$ or from $0$ to a positive real
number R, defining the IRT parameters and proposing an empirical anchor
procedure. As the academic abilities and the school bullying victimization can
be explained due to associated factors such as habits, sex, socioeconomic level
and education level of parents, IRT regression models are proposed to make
joint inferences about individual and school characteristic effects. Results
from the application of the proposed models to the Bogot\'a school bullying
dataset are presented. The need for testing based in statistical models
increases in different fields.
| stat |
Improving Differentially Private Models with Active Learning | Broad adoption of machine learning techniques has increased privacy concerns
for models trained on sensitive data such as medical records. Existing
techniques for training differentially private (DP) models give rigorous
privacy guarantees, but applying these techniques to neural networks can
severely degrade model performance. This performance reduction is an obstacle
to deploying private models in the real world. In this work, we improve the
performance of DP models by fine-tuning them through active learning on public
data. We introduce two new techniques - DIVERSEPUBLIC and NEARPRIVATE - for
doing this fine-tuning in a privacy-aware way. For the MNIST and SVHN datasets,
these techniques improve state-of-the-art accuracy for DP models while
retaining privacy guarantees.
| stat |
Adversarial Robustness as a Prior for Learned Representations | An important goal in deep learning is to learn versatile, high-level feature
representations of input data. However, standard networks' representations seem
to possess shortcomings that, as we illustrate, prevent them from fully
realizing this goal. In this work, we show that robust optimization can be
re-cast as a tool for enforcing priors on the features learned by deep neural
networks. It turns out that representations learned by robust models address
the aforementioned shortcomings and make significant progress towards learning
a high-level encoding of inputs. In particular, these representations are
approximately invertible, while allowing for direct visualization and
manipulation of salient input features. More broadly, our results indicate
adversarial robustness as a promising avenue for improving learned
representations. Our code and models for reproducing these results is available
at https://git.io/robust-reps .
| stat |
Understanding Player Engagement and In-Game Purchasing Behavior with
Ensemble Learning | As video games attract more and more players, the major challenge for game
studios is to retain them. We present a deep behavioral analysis of churn (game
abandonment) and what we called "purchase churn" (the transition from paying to
non-paying user). A series of churning behavior profiles are identified, which
allows a classification of churners in terms of whether they eventually return
to the game (false churners)--or start purchasing again (false purchase
churners)--and their subsequent behavior. The impact of excluding some or all
of these churners from the training sample is then explored in several churn
and purchase churn prediction models. Our results suggest that discarding
certain combinations of "zombies" (players whose activity is extremely
sporadic) and false churners has a significant positive impact in all models
considered.
| stat |
Efficient Bernoulli factory MCMC for intractable posteriors | Accept-reject based Markov chain Monte Carlo (MCMC) algorithms have
traditionally utilised acceptance probabilities that can be explicitly written
as a function of the ratio of the target density at the two contested points.
This feature is rendered almost useless in Bayesian posteriors with unknown
functional forms. We introduce a new family of MCMC acceptance probabilities
that has the distinguishing feature of not being a function of the ratio of the
target density at the two points. We present two stable Bernoulli factories
that generate events within this class of acceptance probabilities. The
efficiency of our methods rely on obtaining reasonable local upper or lower
bounds on the target density and we present two classes of problems where such
bounds are viable: Bayesian inference for diffusions and MCMC on constrained
spaces. The resulting portkey Barker's algorithms are exact and computationally
more efficient that the current state-of-the-art.
| stat |
A Bottom-up Approach to Testing Hypotheses That Have a Branching Tree
Dependence Structure, with False Discovery Rate Control | Modern statistical analyses often involve testing large numbers of
hypotheses. In many situations, these hypotheses may have an underlying tree
structure that not only helps determine the order that tests should be
conducted but also imposes a dependency between tests that must be accounted
for. Our motivating example comes from testing the association between a trait
of interest and groups of microbes that have been organized into operational
taxonomic units (OTUs) or amplicon sequence variants (ASVs). Given p-values
from association tests for each individual OTU or ASV, we would like to know if
we can declare that a certain species, genus, or higher taxonomic grouping can
be considered to be associated with the trait. For this problem, a bottom-up
testing algorithm that starts at the lowest level of the tree (OTUs or ASVs)
and proceeds upward through successively higher taxonomic groupings (species,
genus, family etc.) is required. We develop such a bottom-up testing algorithm
that controls the error rate of decisions made at higher levels in the tree,
conditional on findings at lower levels in the tree. We further show this
algorithm controls the false discovery rate based on the global null hypothesis
that no taxa are associated with the trait. By simulation, we also show that
our approach is better at finding driver taxa, the highest level taxa below
which there are dense association signals. We illustrate our approach using
data from a study of the microbiome among patients with ulcerative colitis and
healthy controls.
| stat |
Beyond Scalar Treatment: A Causal Analysis of Hippocampal Atrophy on
Behavioral Deficits in Alzheimer's Studies | Alzheimer's disease is a progressive form of dementia that results in
problems with memory, thinking and behavior. It often starts with abnormal
aggregation and deposition of beta-amyloid and tau, followed by neuronal damage
such as atrophy of the hippocampi, and finally leads to behavioral deficits.
Despite significant progress in finding biomarkers associated with behavioral
deficits, the underlying causal mechanism remains largely unknown. Here we
investigate whether and how hippocampal atrophy contributes to behavioral
deficits based on a large-scale observational study conducted by the
Alzheimer's Disease Neuroimaging Initiative (ADNI). As a key novelty, we use 2D
representations of the hippocampi, which allows us to better understand atrophy
associated with different subregions. It, however, introduces methodological
challenges as existing causal inference methods are not well suited for
exploiting structural information embedded in the 2D exposures. Moreover, our
data contain more than 6 million clinical and genetic covariates, necessitating
appropriate confounder selection methods. We hence develop a novel two-step
causal inference approach tailored for our ADNI data application. Analysis
results suggest that atrophy of CA1 and subiculum subregions may cause more
severe behavioral deficits compared to CA2 and CA3 subregions. We further
evaluate our method using simulations and provide theoretical guarantees.
| stat |
On the Impact of the Activation Function on Deep Neural Networks
Training | The weight initialization and the activation function of deep neural networks
have a crucial impact on the performance of the training procedure. An
inappropriate selection can lead to the loss of information of the input during
forward propagation and the exponential vanishing/exploding of gradients during
back-propagation. Understanding the theoretical properties of untrained random
networks is key to identifying which deep networks may be trained successfully
as recently demonstrated by Samuel et al (2017) who showed that for deep
feedforward neural networks only a specific choice of hyperparameters known as
the `Edge of Chaos' can lead to good performance. While the work by Samuel et
al (2017) discuss trainability issues, we focus here on training acceleration
and overall performance. We give a comprehensive theoretical analysis of the
Edge of Chaos and show that we can indeed tune the initialization parameters
and the activation function in order to accelerate the training and improve the
performance.
| stat |
Distributed Bootstrap for Simultaneous Inference Under High
Dimensionality | We propose a distributed bootstrap method for simultaneous inference on
high-dimensional massive data that are stored and processed with many machines.
The method produces a $\ell_\infty$-norm confidence region based on a
communication-efficient de-biased lasso, and we propose an efficient
cross-validation approach to tune the method at every iteration. We
theoretically prove a lower bound on the number of communication rounds
$\tau_{\min}$ that warrants the statistical accuracy and efficiency.
Furthermore, $\tau_{\min}$ only increases logarithmically with the number of
workers and intrinsic dimensionality, while nearly invariant to the nominal
dimensionality. We test our theory by extensive simulation studies, and a
variable screening task on a semi-synthetic dataset based on the US Airline
On-time Performance dataset. The code to reproduce the numerical results is
available at GitHub: https://github.com/skchao74/Distributed-bootstrap.
| stat |
Group Decisions based on Confidence Weighted Majority Voting | Background: It has repeatedly been reported that when making decisions under
uncertainty, groups outperform individuals. In a lab setting, real groups are
often replaced by simulated groups: Instead of performing an actual group
discussion, individual responses are aggregated by a numerical computation.
While studies typically use unweighted majority voting (MV) for this
aggregation, the theoretically optimal method is confidence weighted majority
voting (CWMV) - if confidence ratings for the individual responses are
available. However, it is not entirely clear how well the theoretically derived
CWMV method predicts real group decisions and confidences. Therefore, we
compared simulated group responses using CWMV and MV to real group responses.
Results: Simulated group decisions based on CWMV matched the accuracy of real
group decisions very well, while simulated group decisions based on MV showed
lower accuracy. Also, CWMV well predicted the confidence that groups put into
their group decisions. Yet, individuals and real groups showed a bias towards
under--confidence while CWMV does not.
Conclusion: Our results highlight the importance of taking into account
individual confidences when investigating group decisions: We found that real
groups can aggregate individual confidences such that they match the optimal
aggregation given by CWMV. This implies that research using simulated group
decisions should use CWMV and not MV.
| stat |
Multiple Target, Multiple Type Filtering in the RFS Framework | A Multiple Target, Multiple Type Filtering (MTMTF) algorithm is developed
using Random Finite Set (RFS) theory. First, we extend the standard Probability
Hypothesis Density (PHD) filter for multiple types of targets, each with
distinct detection properties, to develop a multiple target, multiple type
filtering, N-type PHD filter, where $N\geq2$, for handling confusions among
target types. In this approach, we assume that there will be confusions between
detections, i.e. clutter arises not just from background false positives, but
also from target confusions. Then, under the assumptions of Gaussianity and
linearity, we extend the Gaussian mixture (GM) implementation of the standard
PHD filter for the proposed N-type PHD filter termed the N-type GM-PHD filter.
Furthermore, we analyze the results from simulations to track sixteen targets
of four different types using a four-type (quad) GM-PHD filter as a typical
example and compare it with four independent GM-PHD filters using the Optimal
Subpattern Assignment (OSPA) metric. This shows the improved performance of our
strategy that accounts for target confusions by efficiently discriminating
them.
| stat |
Deciphering hierarchical organization of topologically associated
domains through change-point testing | Background: The nucleus of eukaryotic cells spatially packages chromosomes
into a hierarchical and distinct segregation that plays critical roles in
maintaining transcription regulation. High-throughput methods of chromosome
conformation capture, such as Hi-C, have revealed topologically associating
domains (TADs) that are defined by biased chromatin interactions within them.
Results: Here, we introduce a novel method, HiCKey, to decipher hierarchical
TAD structures in Hi-C data and compare them across samples. We first derive a
generalized likelihood-ratio (GLR) test for detecting change-points in an
interaction matrix that follows a negative binomial distribution or general
mixture distribution. We then employ several optimal search strategies to
decipher hierarchical TADs with p-values calculated by the GLR test.
Large-scale validations of simulation data show that HiCKey has good precision
in recalling known TADs and is robust against random collision noise of
chromatin interactions. By applying HiCKey to Hi-C data of seven human cell
lines, we identified multiple layers of TAD organization among them, but the
vast majority had no more than four layers. In particular, we found that TAD
boundaries are significantly enriched in active chromosomal regions compared to
repressed regions, indicating finer hierarchical architectures in active
regions for precise gene transcription regulation.
Conclusions: HiCKey is optimized for processing large matrices constructed
from high-resolution Hi-C experiments. The method and theoretical result of the
GLR test provide a general framework for significance testing of similar
experimental chromatin interaction data that may not fully follow negative
binomial distributions but rather more general mixture distributions.
| stat |
A constrained minimum criterion for model selection | We propose a hypothesis test based model selection criterion for the best
subset selection of sparse linear models. We show it is consistent in that the
probability of its choosing the true model approaches one and the parameter
values of its chosen model converge in probability to that of the true model as
the sample size goes to infinity. This criterion is capable of controlling the
balance between the false active rate and false inactive rate of the selected
model, and it can be applied with other methods of model selection such as the
lasso. We also demonstrate its accuracy and advantages with a numerical
comparison and an application.
| stat |
Active Sampling for the Quickest Detection of Markov Networks | Consider $n$ random variables forming a Markov random field (MRF). The true
model of the MRF is unknown, and it is assumed to belong to a binary set. The
objective is to sequentially sample the random variables (one-at-a-time) such
that the true MRF model can be detected with the fewest number of samples,
while in parallel, the decision reliability is controlled. The core element of
an optimal decision process is a rule for selecting and sampling the random
variables over time. Such a process, at every time instant and adaptively to
the collected data, selects the random variable that is expected to be most
informative about the model, rendering an overall minimized number of samples
required for reaching a reliable decision. The existing studies on detecting
MRF structures generally sample the entire network at the same time and focus
on designing optimal detection rules without regard to the data-acquisition
process. This paper characterizes the sampling process for general MRFs, which,
in conjunction with the sequential probability ratio test, is shown to be
optimal in the asymptote of large $n$. The critical insight in designing the
sampling process is devising an information measure that captures the
decisions' inherent statistical dependence over time. Furthermore, when the
MRFs can be modeled by acyclic probabilistic graphical models, the sampling
rule is shown to take a computationally simple form. Performance analysis for
the general case is provided, and the results are interpreted in several
special cases: Gaussian MRFs, non-asymptotic regimes, connection to Chernoff's
rule to controlled (active) sensing, and the problem of cluster detection.
| stat |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.