title
stringlengths 6
244
| abstract
stringlengths 19
5.09k
| label
stringclasses 10
values |
|---|---|---|
Multi-Output Gaussian Processes for Multi-Population Longevity Modeling | We investigate joint modeling of longevity trends using the spatial
statistical framework of Gaussian Process regression. Our analysis is motivated
by the Human Mortality Database (HMD) that provides unified raw mortality
tables for nearly 40 countries. Yet few stochastic models exist for handling
more than two populations at a time. To bridge this gap, we leverage a spatial
covariance framework from machine learning that treats populations as distinct
levels of a factor covariate, explicitly capturing the cross-population
dependence. The proposed multi-output Gaussian Process models straightforwardly
scale up to a dozen populations and moreover intrinsically generate coherent
joint longevity scenarios. In our numerous case studies we investigate
predictive gains from aggregating mortality experience across nations and
genders, including by borrowing the most recently available "foreign" data. We
show that in our approach, information fusion leads to more precise (and
statistically more credible) forecasts. We implement our models in \texttt{R},
as well as a Bayesian version in \texttt{Stan} that provides further
uncertainty quantification regarding the estimated mortality covariance
structure. All examples utilize public HMD datasets.
| stat |
Survey of Imbalanced Data Methodologies | Imbalanced data set is a problem often found and well-studied in financial
industry. In this paper, we reviewed and compared some popular methodologies
handling data imbalance. We then applied the under-sampling/over-sampling
methodologies to several modeling algorithms on UCI and Keel data sets. The
performance was analyzed for class-imbalance methods, modeling algorithms and
grid search criteria comparison.
| stat |
On parameter estimation with the Wasserstein distance | Statistical inference can be performed by minimizing, over the parameter
space, the Wasserstein distance between model distributions and the empirical
distribution of the data. We study asymptotic properties of such minimum
Wasserstein distance estimators, complementing results derived by Bassetti,
Bodini and Regazzini in 2006. In particular, our results cover the misspecified
setting, in which the data-generating process is not assumed to be part of the
family of distributions described by the model. Our results are motivated by
recent applications of minimum Wasserstein estimators to complex generative
models. We discuss some difficulties arising in the approximation of these
estimators and illustrate their behavior in several numerical experiments. Two
of our examples are taken from the literature on approximate Bayesian
computation and have likelihood functions that are not analytically tractable.
Two other examples involve misspecified models.
| stat |
Parsimonious Feature Extraction Methods: Extending Robust Probabilistic
Projections with Generalized Skew-t | We propose a novel generalisation to the Student-t Probabilistic Principal
Component methodology which: (1) accounts for an asymmetric distribution of the
observation data; (2) is a framework for grouped and generalised
multiple-degree-of-freedom structures, which provides a more flexible approach
to modelling groups of marginal tail dependence in the observation data; and
(3) separates the tail effect of the error terms and factors. The new feature
extraction methods are derived in an incomplete data setting to efficiently
handle the presence of missing values in the observation vector. We discuss
various special cases of the algorithm being a result of simplified assumptions
on the process generating the data. The applicability of the new framework is
illustrated on a data set that consists of crypto currencies with the highest
market capitalisation.
| stat |
Surrogate-based Optimization using Mutual Information for Computer
Experiments (optim-MICE) | The computational burden of running a complex computer model can make
optimization impractical. Gaussian Processes (GPs) are statistical surrogates
(also known as emulators) that alleviate this issue since they cheaply replace
the computer model. As a result, the exploration vs. exploitation trade-off
strategy can be accelerated by building a GP surrogate. In this paper, we
propose a new surrogate-based optimization scheme that minimizes the number of
evaluations of the computationally expensive function. Taking advantage of
parallelism of the evaluation of the unknown function, the uncertain regions
are explored simultaneously, and a batch of input points is chosen using Mutual
Information for Computer Experiments (MICE), a sequential design algorithm
which maximizes the information theoretic Mutual Information over the input
space. The computational efficiency of interweaving the optimization scheme
with MICE (optim-MICE) is examined and demonstrated on test functions.
Optim-MICE is compared with state-of-the-art heuristics such as Efficient
Global Optimization (EGO) and GP-Upper Confidence Bound (GP-UCB). We
demonstrate that optim-MICE outperforms these schemes on a large range of
computational experiments.
| stat |
Trial emulation and survival analysis for disease incidence registers: a
case study on the causal effect of pre-emptive kidney transplantation | Numerous tutorials and research papers focus on methods in either survival
analysis or causal inference, leaving common complications in medical studies
unaddressed. In practice one must handle problems jointly, without the luxury
of ignoring essential features of the data structure. In this paper, we follow
incident cases of end-stage renal disease and examine the effect on all-cause
mortality of starting treatment with transplant, so-called pre-emptive kidney
transplantation, versus dialysis. The question is relatively simple: which
treatment start is expected to bring the best survival for a target population?
To address the question, we emulate a target trial drawing on the Swedish Renal
Registry to estimate a causal effect on survival curves. Aware of important
challenges, we see how previous studies have selected patients into treatment
groups based on events occurring post treatment initiation. Our study reveals
the dramatic impact of resulting immortal time bias and other typical features
of long term incident disease registries, including: missing or mismeasured
covariates during (the early) phases of the register, varying risk profile of
patients entering treatment groups over calendar time and changes in risk as
care improves over the years. With characteristics of cases and versions of
treatment evolving over time, informative censoring is introduced in unadjusted
Kaplan-Meier curves and also their IPW version is no longer valid. Here we
discuss feasible ways of handling these features and answer different research
questions relying on the no unmeasured baseline confounders assumption.
| stat |
Doubly robust capture-recapture methods for estimating population size | Estimation of population size using incomplete lists (also called the
capture-recapture problem) has a long history across many biological and social
sciences. For example, human rights and other groups often construct partial
and overlapping lists of victims of armed conflicts, with the hope of using
this information to estimate the total number of victims. Earlier statistical
methods for this setup either use potentially restrictive parametric
assumptions, or else rely on typically suboptimal plug-in-type nonparametric
estimators; however, both approaches can lead to substantial bias, the former
via model misspecification and the latter via smoothing. Under an identifying
assumption that two lists are conditionally independent given measured
covariate information, we make several contributions. First we derive the
nonparametric efficiency bound for estimating the capture probability, which
indicates the best possible performance of any estimator, and sheds light on
the statistical limits of capture-recapture methods. Then we present a new
estimator, and study its finite-sample properties, showing that it has a double
robustness property new to capture-recapture, and that it is near-optimal in a
non-asymptotic sense, under relatively mild nonparametric conditions. Next, we
give a method for constructing confidence intervals for total population size
from generic capture probability estimators, and prove non-asymptotic
near-validity. Finally, we study our methods in simulations, and apply them to
estimate the number of killings and disappearances attributable to different
groups in Peru during its internal armed conflict between 1980 and 2000.
| stat |
Unsupervised Learning with Stein's Unbiased Risk Estimator | Learning from unlabeled and noisy data is one of the grand challenges of
machine learning. As such, it has seen a flurry of research with new ideas
proposed continuously. In this work, we revisit a classical idea: Stein's
Unbiased Risk Estimator (SURE). We show that, in the context of image recovery,
SURE and its generalizations can be used to train convolutional neural networks
(CNNs) for a range of image denoising and recovery problems without any ground
truth data.
Specifically, our goal is to reconstruct an image $x$ from a noisy linear
transformation (measurement) of the image. We consider two scenarios: one where
no additional data is available and one where we have measurements of other
images that are drawn from the same noisy distribution as $x$, but have no
access to the clean images. Such is the case, for instance, in the context of
medical imaging, microscopy, and astronomy, where noise-less ground truth data
is rarely available.
We show that in this situation, SURE can be used to estimate the
mean-squared-error loss associated with an estimate of $x$. Using this estimate
of the loss, we train networks to perform denoising and compressed sensing
recovery. In addition, we also use the SURE framework to partially explain and
improve upon an intriguing results presented by Ulyanov et al. in "Deep Image
Prior": that a network initialized with random weights and fit to a single
noisy image can effectively denoise that image.
Public implementations of the networks and methods described in this paper
can be found at https://github.com/ricedsp/D-AMP_Toolbox.
| stat |
Generalized Variational Inference: Three arguments for deriving new
Posteriors | We advocate an optimization-centric view on and introduce a novel
generalization of Bayesian inference. Our inspiration is the representation of
Bayes' rule as infinite-dimensional optimization problem (Csiszar, 1975;
Donsker and Varadhan; 1975, Zellner; 1988). First, we use it to prove an
optimality result of standard Variational Inference (VI): Under the proposed
view, the standard Evidence Lower Bound (ELBO) maximizing VI posterior is
preferable to alternative approximations of the Bayesian posterior. Next, we
argue for generalizing standard Bayesian inference. The need for this arises in
situations of severe misalignment between reality and three assumptions
underlying standard Bayesian inference: (1) Well-specified priors, (2)
well-specified likelihoods, (3) the availability of infinite computing power.
Our generalization addresses these shortcomings with three arguments and is
called the Rule of Three (RoT). We derive it axiomatically and recover existing
posteriors as special cases, including the Bayesian posterior and its
approximation by standard VI. In contrast, approximations based on alternative
ELBO-like objectives violate the axioms. Finally, we study a special case of
the RoT that we call Generalized Variational Inference (GVI). GVI posteriors
are a large and tractable family of belief distributions specified by three
arguments: A loss, a divergence and a variational family. GVI posteriors have
appealing properties, including consistency and an interpretation as
approximate ELBO. The last part of the paper explores some attractive
applications of GVI in popular machine learning models, including robustness
and more appropriate marginals. After deriving black box inference schemes for
GVI posteriors, their predictive performance is investigated on Bayesian Neural
Networks and Deep Gaussian Processes, where GVI can comprehensively improve
upon existing methods.
| stat |
Distribution-free calibration guarantees for histogram binning without
sample splitting | We prove calibration guarantees for the popular histogram binning (also
called uniform-mass binning) method of Zadrozny and Elkan [2001]. Histogram
binning has displayed strong practical performance, but theoretical guarantees
have only been shown for sample split versions that avoid 'double dipping' the
data. We demonstrate that the statistical cost of sample splitting is
practically significant on a credit default dataset. We then prove calibration
guarantees for the original method that double dips the data, using a certain
Markov property of order statistics. Based on our results, we make practical
recommendations for choosing the number of bins in histogram binning. In our
illustrative simulations, we propose a new tool for assessing calibration --
validity plots -- which provide more information than an ECE estimate.
| stat |
Estimating heterogeneous survival treatment effect in observational data
using machine learning | Methods for estimating heterogeneous treatment effect in observational data
have largely focused on continuous or binary outcomes, and have been relatively
less vetted with survival outcomes. Using flexible machine learning methods in
the counterfactual framework is a promising approach to address challenges due
to complex individual characteristics, to which treatments need to be tailored.
To evaluate the operating characteristics of recent survival machine learning
methods for the estimation of treatment effect heterogeneity and inform better
practice, we carry out a comprehensive simulation study presenting a wide range
of settings describing confounded heterogeneous survival treatment effects and
varying degrees of covariate overlap. Our results suggest that the
nonparametric Bayesian Additive Regression Trees within the framework of
accelerated failure time model (AFT-BART-NP) consistently yields the best
performance, in terms of bias, precision and expected regret. Moreover, the
credible interval estimators from AFT-BART-NP provide close to nominal
frequentist coverage for the individual survival treatment effect when the
covariate overlap is at least moderate. Including a non-parametrically
estimated propensity score as an additional fixed covariate in the AFT-BART-NP
model formulation can further improve its efficiency and frequentist coverage.
Finally, we demonstrate the application of flexible causal machine learning
estimators through a comprehensive case study examining the heterogeneous
survival effects of two radiotherapy approaches for localized high-risk
prostate cancer.
| stat |
Bayesian Auxiliary Variable Model for Birth Records Data with
Qualitative and Quantitative Responses | Many applications involve data with qualitative and quantitative responses.
When there is an association between the two responses, a joint model will
provide improved results than modeling them separately. In this paper, we
propose a Bayesian method to jointly model such data. The joint model links the
qualitative and quantitative responses and can assess their dependency strength
via a latent variable. The posterior distributions of parameters are obtained
through an efficient MCMC sampling algorithm. The simulation shows that the
proposed method can improve the prediction capacity for both responses. We
apply the proposed joint model to the birth records data acquired by the
Virginia Department of Health and study the mutual dependence between preterm
birth of infants and their birth weights.
| stat |
Bayesian Hierarchical Models with Conjugate Full-Conditional
Distributions for Dependent Data from the Natural Exponential Family | We introduce a Bayesian approach for analyzing (possibly) high-dimensional
dependent data that are distributed according to a member from the natural
exponential family of distributions. This problem requires extensive
methodological advancements, as jointly modeling high-dimensional dependent
data leads to the so-called "big n problem." The computational complexity of
the "big n problem" is further exacerbated when allowing for non-Gaussian data
models, as is the case here. Thus, we develop new computationally efficient
distribution theory for this setting. In particular, we introduce the
"conjugate multivariate distribution," which is motivated by the univariate
distribution introduced in Diaconis and Ylvisaker (1979). Furthermore, we
provide substantial theoretical and methodological development including:
results regarding conditional distributions, an asymptotic relationship with
the multivariate normal distribution, conjugate prior distributions, and
full-conditional distributions for a Gibbs sampler. To demonstrate the
wide-applicability of the proposed methodology, we provide two simulation
studies and three applications based on an epidemiology dataset, a federal
statistics dataset, and an environmental dataset, respectively.
| stat |
Machine Learning Assisted Orthonormal Basis Selection for Functional
Data Analysis | In implementations of the functional data methods, the effect of the initial
choice of an orthonormal basis has not gained much attention in the past.
Typically, several standard bases such as Fourier, wavelets, splines, etc. are
considered to transform observed functional data and a choice is made without
any formal criteria indicating which of the bases is preferable for the initial
transformation of the data into functions. In an attempt to address this issue,
we propose a strictly data-driven method of orthogonal basis selection. The
method uses recently introduced orthogonal spline bases called the splinets
obtained by efficient orthogonalization of the B-splines. The algorithm learns
from the data in the machine learning style to efficiently place knots. The
optimality criterion is based on the average (per functional data point) mean
square error and is utilized both in the learning algorithms and in comparison
studies. The latter indicates efficiency that is particularly evident for the
sparse functional data and to a lesser degree in analyses of responses to
complex physical systems.
| stat |
Automated Vehicle Crash Sequences: Patterns and Potential Uses in Safety
Testing | With safety being one of the primary motivations for developing automated
vehicles (AVs), extensive field and simulation tests are being carried out to
ensure AVs can operate safely on roadways. Since 2014, the California DMV has
been collecting AV collision and disengagement reports, which are valuable data
sources for studying AV crash patterns. In this study, crash sequence data
extracted from California AV collision reports were used to investigate
patterns and how they may be used to develop AV test scenarios. Employing
sequence analysis, this study evaluated 168 AV crashes (with AV in automatic
driving mode before disengagement or collision) from 2015 to 2019. Analysis of
subsequences showed that the most representative pattern in AV crashes was
(collision following AV stop) type. Analysis of event transition showed that
disengagement, as an event in 24 percent of all studied AV crash sequences, had
a transition probability of 68 percent to an immediate collision. Cluster
analysis characterized AV crash sequences into seven groups with distinctive
crash dynamic features. Cross-tabulation analysis showed that sequence groups
were significantly associated with variables measuring crash outcomes and
describing environmental conditions. Crash sequences are useful for developing
AV test scenarios. Based on the findings, a scenario-based AV safety testing
framework was proposed with sequence of events embedded as a core component.
| stat |
Employing latent variable models to improve efficiency in composite
endpoint analysis | Composite endpoints that combine multiple outcomes on different scales are
common in clinical trials, particularly in chronic conditions. In many of these
cases, patients will have to cross a predefined responder threshold in each of
the outcomes to be classed as a responder overall. One instance of this occurs
in systemic lupus erythematosus (SLE), where the responder endpoint combines
two continuous, one ordinal and one binary measure. The overall binary
responder endpoint is typically analysed using logistic regression, resulting
in a substantial loss of information. We propose a latent variable model for
the SLE endpoint, which assumes that the discrete outcomes are manifestations
of latent continuous measures and can proceed to jointly model the components
of the composite. We perform a simulation study and find the method to offer
large efficiency gains over the standard analysis. We find that the magnitude
of the precision gains are highly dependent on which components are driving
response. Bias is introduced when joint normality assumptions are not
satisfied, which we correct for using a bootstrap procedure. The method is
applied to the Phase IIb MUSE trial in patients with moderate to severe SLE. We
show that it estimates the treatment effect 2.5 times more precisely, offering
a 60% reduction in required sample size.
| stat |
The Locally Gaussian Partial Correlation | It is well known that the dependence structure for jointly Gaussian variables
can be fully captured using correlations, and that the conditional dependence
structure in the same way can be described using partial correlations. The
partial orrelation does not, however, characterize conditional dependence in
many non-Gaussian populations. This paper introduces the local Gaussian partial
correlation (LGPC), a new measure of conditional dependence. It is a local
version of the partial correlation coefficient that characterizes conditional
dependence in a large class of populations. It has some useful and novel
properties besides: The LGPC reduces to the ordinary partial correlation for
jointly normal variables, and it distinguishes between positive and negative
conditional dependence. Furthermore, the LGPC can be used to study departures
from conditional independence in specific parts of the distribution. We provide
several examples of this, both simulated and real, and derive estimation theory
under a local likelihood framework. Finally, we indicate how the LGPC can be
used to construct a powerful test for conditional independence, which, again,
can be used to detect Granger causality in time series.
| stat |
Learning Structure in Nested Logit Models | This paper introduces a new data-driven methodology for nested logit
structure discovery. Nested logit models allow the modeling of positive
correlations between the error terms of the utility specifications of the
different alternatives in a discrete choice scenario through the specification
of a nesting structure. Current nested logit model estimation practices require
an a priori specification of a nesting structure by the modeler. In this we
work we optimize over all possible specifications of the nested logit model
that are consistent with rational utility maximization. We formulate the
problem of learning an optimal nesting structure from the data as a mixed
integer nonlinear programming (MINLP) optimization problem and solve it using a
variant of the linear outer approximation algorithm. We exploit the tree
structure of the problem and utilize the latest advances in integer
optimization to bring practical tractability to the optimization problem we
introduce. We demonstrate the ability of our algorithm to correctly recover the
true nesting structure from synthetic data in a Monte Carlo experiment. In an
empirical illustration using a stated preference survey on modes of
transportation in the U.S. state of Massachusetts, we use our algorithm to
obtain an optimal nesting tree representing the correlations between the
unobserved effects of the different travel mode choices. We provide our
implementation as a customizable and open-source code base written in the Julia
programming language.
| stat |
National mapping and estimation of forest area by dominant tree species
using Sentinel-2 data | Nation-wide Sentinel-2 mosaics were used with National Forest Inventory (NFI)
data for modelling and subsequent mapping of spruce, pine and deciduous forest
in Norway in 16 m $\times$ 16 m resolution. The accuracies of the best model
ranged between 74% for spruce and 87% for deciduous forest. An overall accuracy
of 90% was found on stand level using independent data from more than 42.000
stands. Errors mostly resulting from a forest mask reduced the model accuracies
by approximately 10%. Nonetheless, efficiencies of national forest area
estimates increased by 20% to 50% for model-assisted (MA) estimates and up to
90% for poststratification (PS). Greater minimum number of observations,
however, constrained the use of PS. For estimates of sub-populations
(municipalities), efficiencies improved by up to a factor of 8. However, even
for municipalities with a decent number of NFI plots, direct NFI estimates were
sometimes more precise than MA estimates. PS estimates were always more precise
than direct and MA estimates but were applicable in fewer municipalities. The
tree species prediction map is freely available as part of the Norwegian forest
resource map and is used, among others, to improve maps of other variables of
interest such as timber volume and biomass.
| stat |
Forecasting COVID-19 daily cases using phone call data | The need to forecast COVID-19 related variables continues to be pressing as
the epidemic unfolds. Different efforts have been made, with compartmental
models in epidemiology and statistical models such as AutoRegressive Integrated
Moving Average (ARIMA), Exponential Smoothing (ETS) or computing intelligence
models. These efforts have proved useful in some instances by allowing decision
makers to distinguish different scenarios during the emergency, but their
accuracy has been disappointing, forecasts ignore uncertainties and less
attention is given to local areas. In this study, we propose a simple Multiple
Linear Regression model, optimised to use call data to forecast the number of
daily confirmed cases. Moreover, we produce a probabilistic forecast that
allows decision makers to better deal with risk. Our proposed approach
outperforms ARIMA, ETS and a regression model without call data, evaluated by
three point forecast error metrics, one prediction interval and two
probabilistic forecast accuracy measures. The simplicity, interpretability and
reliability of the model, obtained in a careful forecasting exercise, is a
meaningful contribution to decision makers at local level who acutely need to
organise resources in already strained health services. We hope that this model
would serve as a building block of other forecasting efforts that on the one
hand would help front-line personal and decision makers at local level, and on
the other would facilitate the communication with other modelling efforts being
made at the national level to improve the way we tackle this pandemic and other
similar future challenges.
| stat |
Bayesian dose-regimen assessment in early phase oncology incorporating
pharmacokinetics and pharmacodynamics | Phase I dose-finding trials in oncology seek to find the maximum tolerated
dose (MTD) of a drug under a specific schedule. Evaluating drug-schedules aims
at improving treatment safety while maintaining efficacy. However, while we can
reasonably assume that toxicity increases with the dose for cytotoxic drugs,
the relationship between toxicity and multiple schedules remains elusive. We
proposed a Bayesian dose-regimen assessment method (DRtox) using
pharmacokinetics/pharmacodynamics (PK/PD) information to estimate the maximum
tolerated dose-regimen (MTD-regimen), at the end of the dose-escalation stage
of a trial to be recommended for the next phase. We modeled the binary toxicity
via a PD endpoint and estimated the dose-regimen toxicity relationship through
the integration of a dose-regimen PD model and a PD toxicity model. For the
dose-regimen PD model, we considered nonlinear mixed-effects models, and for
the PD toxicity model, we proposed the following two Bayesian approaches: a
logistic model and a hierarchical model. We evaluated the operating
characteristics of the DRtox through simulation studies under various
scenarios. The results showed that our method outperforms traditional
model-based designs demonstrating a higher percentage of correctly selecting
the MTD-regimen. Moreover, the inclusion of PK/PD information in the DRtox
helped provide more precise estimates for the entire dose-regimen toxicity
curve; therefore the DRtox may recommend alternative untested regimens for
expansion cohorts. The DRtox should be applied at the end of the
dose-escalation stage of an ongoing trial for patients with relapsed or
refractory acute myeloid leukemia (NCT03594955) once all toxicity and PK/PD
data are collected.
| stat |
A Hierarchical Spatio-Temporal Statistical Model Motivated by Glaciology | In this paper, we extend and analyze a Bayesian hierarchical spatio-temporal
model for physical systems. A novelty is to model the discrepancy between the
output of a computer simulator for a physical process and the actual process
values with a multivariate random walk. For computational efficiency, linear
algebra for bandwidth limited matrices is utilized, and first-order emulator
inference allows for the fast emulation of a numerical partial differential
equation (PDE) solver. A test scenario from a physical system motivated by
glaciology is used to examine the speed and accuracy of the computational
methods used, in addition to the viability of modeling assumptions. We conclude
by discussing how the model and associated methodology can be applied in other
physical contexts besides glaciology.
| stat |
Confidence Tubes for Curves on SO(3) and Identification of
Subject-Specific Gait Change after Kneeling | In order to identify changes of gait patterns, e.g. due to prolonged
occupational kneeling, which is believed to be major risk factor, among others,
for the development of knee osteoarthritis, we develop confidence tubes for
curves following a Gaussian perturbation model on SO(3). These are based on an
application of the Gaussian kinematic formula to a process of Hotelling
statistics and we approximate them by a computible version, for which we show
convergence. Simulations endorse our method, which in application to gait
curves from eight volunteers undergoing kneeling tasks, identifies phases of
the gait cycle that have changed due to kneeling tasks. We find that after
kneeling, deviation from normal gait is stronger, in particular for older aged
male volunteers. Notably our method adjusts for different walking speeds and
marker replacement at different visits.
| stat |
Gibbs flow for approximate transport with applications to Bayesian
computation | Let $\pi_{0}$ and $\pi_{1}$ be two distributions on the Borel space
$(\mathbb{R}^{d},\mathcal{B}(\mathbb{R}^{d}))$. Any measurable function
$T:\mathbb{R}^{d}\rightarrow\mathbb{R}^{d}$ such that $Y=T(X)\sim\pi_{1}$ if
$X\sim\pi_{0}$ is called a transport map from $\pi_{0}$ to $\pi_{1}$. For any
$\pi_{0}$ and $\pi_{1}$, if one could obtain an analytical expression for a
transport map from $\pi_{0}$ to $\pi_{1}$, then this could be straightforwardly
applied to sample from any distribution. One would map draws from an
easy-to-sample distribution $\pi_{0}$ to the target distribution $\pi_{1}$
using this transport map. Although it is usually impossible to obtain an
explicit transport map for complex target distributions, we show here how to
build a tractable approximation of a novel transport map. This is achieved by
moving samples from $\pi_{0}$ using an ordinary differential equation with a
velocity field that depends on the full conditional distributions of the
target. Even when this ordinary differential equation is time-discretized and
the full conditional distributions are numerically approximated, the resulting
distribution of mapped samples can be efficiently evaluated and used as a
proposal within sequential Monte Carlo samplers. We demonstrate significant
gains over state-of-the-art sequential Monte Carlo samplers at a fixed
computational complexity on a variety of applications.
| stat |
Predicting feature imputability in the absence of ground truth | Data imputation is the most popular method of dealing with missing values,
but in most real life applications, large missing data can occur and it is
difficult or impossible to evaluate whether data has been imputed accurately
(lack of ground truth). This paper addresses these issues by proposing an
effective and simple principal component based method for determining whether
individual data features can be accurately imputed - feature imputability. In
particular, we establish a strong linear relationship between principal
component loadings and feature imputability, even in the presence of extreme
missingness and lack of ground truth. This work will have important
implications in practical data imputation strategies.
| stat |
Forecasting observables with particle filters: Any filter will do! | We investigate the impact of filter choice on forecast accuracy in state
space models. The filters are used both to estimate the posterior distribution
of the parameters, via a particle marginal Metropolis-Hastings (PMMH)
algorithm, and to produce draws from the filtered distribution of the final
state. Multiple filters are entertained, including two new data-driven methods.
Simulation exercises are used to document the performance of each PMMH
algorithm, in terms of computation time and the efficiency of the chain. We
then produce the forecast distributions for the one-step-ahead value of the
observed variable, using a fixed number of particles and Markov chain draws.
Despite distinct differences in efficiency, the filters yield virtually
identical forecasting accuracy, with this result holding under both correct and
incorrect specification of the model. This invariance of forecast performance
to the specification of the filter also characterizes an empirical analysis of
S&P500 daily returns.
| stat |
Regularized Fingerprinting in Detection and Attribution of Climate
Change with Weight Matrix Optimizing the Efficiency in Scaling Factor
Estimation | The optimal fingerprinting method for detection and attribution of climate
change is based on a multiple regression where each covariate has measurement
error whose covariance matrix is the same as that of the regression error up to
a known scale. Inferences about the regression coefficients are critical not
only for making statements about detection and attribution but also for
quantifying the uncertainty in important outcomes derived from detection and
attribution analyses. When there is no errors-in-variables (EIV), the optimal
weight matrix in estimating the regression coefficients is the precision matrix
of the regression error which, in practice, is never known and has to be
estimated from climate model simulations. We construct a weight matrix by
inverting a nonlinear shrinkage estimate of the error covariance matrix that
minimizes loss functions directly targeting the uncertainty of the resulting
regression coefficient estimator. The resulting estimator of the regression
coefficients is asymptotically optimal as the sample size of the climate model
simulations and the matrix dimension go to infinity together with a limiting
ratio. When EIVs are present, the estimator of the regression coefficients
based on the proposed weight matrix is asymptotically more efficient than that
based on the inverse of the existing linear shrinkage estimator of the error
covariance matrix. The performance of the method is confirmed in finite sample
simulation studies mimicking realistic situations in terms of the length of the
confidence intervals and empirical coverage rates for the regression
coefficients. An application to detection and attribution analyses of the mean
temperature at different spatial scales illustrates the utility of the method.
| stat |
Recursive Formula for Labeled Graph Enumeration | This manuscript presents a general recursive formula to estimate the size of
fibers associated with algebraic maps from graphs to summary statistics of
importance for social network analysis, such as number of edges (graph
density), degree sequence, degree distribution, mixing by nodal covariates, and
degree mixing. That is, the formula estimates the number of labeled graphs that
have given values for network properties. The proposed approach can be extended
to additional network properties (e.g., clustering) as well as properties of
bipartite networks. For special settings in which alternative formulas exist,
simulation studies demonstrate the validity of the proposed approach. We
illustrate the approach for estimating the size of fibers associated with the
Barab\'{a}si--Albert model for the properties of degree distribution and degree
mixing. In addition, we demonstrate how the approach can be used to assess the
diversity of graphs within a fiber.
| stat |
A Novel Algorithm for Optimized Real Time Anomaly Detection in
Timeseries | Observations in data which are significantly different from its neighbouring
points but cannot be classified as noise are known as anomalies or outliers.
These anomalies are a cause of concern and a timely warning about their
presence could be valuable. In this paper, we have evaluated and compared the
performance of popular algorithms from domains of Machine Learning and
Statistics in detecting anomalies on both offline data as well as real time
data. Our aim is to come up with an algorithm which can handle all types of
seasonal and non-seasonal data effectively and is fast enough to be of
practical utility in real time. It is not only important to detect anomalies at
the global but also the ones which are anomalies owing to their local
surroundings. Such outliers can be termed as contextual anomalies as they
derive their context from the neighbouring observations. Also, we require a
methodology to automatically determine the presence of seasonality in the given
data. For detecting the seasonality, the proposed algorithm takes up a curve
fitting approach rather than model based anomaly detection. The proposed model
also introduces a unique filter which assess the relative significance of local
outliers and removes the ones deemed as insignificant. Since, the proposed
model fits polynomial in buckets of timeseries data, it does not suffer from
problems such as heteroskedasticity and breakout as compared to its statistical
alternatives such as ARIMA, SARIMA and Winter Holt. Experimental results the
proposed algorithm performs better on both real time as well as artificial
generated datasets.
| stat |
Automating Involutive MCMC using Probabilistic and Differentiable
Programming | Involutive MCMC is a unifying mathematical construction for MCMC kernels that
generalizes many classic and state-of-the-art MCMC algorithms, from reversible
jump MCMC to kernels based on deep neural networks. But as with MCMC samplers
more generally, implementing involutive MCMC kernels is often tedious and
error-prone, especially when sampling on complex state spaces. This paper
describes a technique for automating the implementation of involutive MCMC
kernels given (i) a pair of probabilistic programs defining the target
distribution and an auxiliary distribution respectively and (ii) a
differentiable program that transforms the execution traces of these
probabilistic programs. The technique, which is implemented as part of the Gen
probabilistic programming system, also automatically detects user errors in the
specification of involutive MCMC kernels and exploits sparsity in the kernels
for improved efficiency. The paper shows example Gen code for a split-merge
reversible jump move in an infinite Gaussian mixture model and a
state-dependent mixture of proposals on a combinatorial space of covariance
functions for a Gaussian process.
| stat |
Factor Analysis of Mixed Data for Anomaly Detection | Anomaly detection aims to identify observations that deviate from the typical
pattern of data. Anomalous observations may correspond to financial fraud,
health risks, or incorrectly measured data in practice. We show detecting
anomalies in high-dimensional mixed data is enhanced through first embedding
the data then assessing an anomaly scoring scheme. We focus on unsupervised
detection and the continuous and categorical (mixed) variable case. We propose
a kurtosis-weighted Factor Analysis of Mixed Data for anomaly detection,
FAMDAD, to obtain a continuous embedding for anomaly scoring. We illustrate
that anomalies are highly separable in the first and last few ordered
dimensions of this space, and test various anomaly scoring experiments within
this subspace. Results are illustrated for both simulated and real datasets,
and the proposed approach (FAMDAD) is highly accurate for high-dimensional
mixed data throughout these diverse scenarios.
| stat |
Continual Learning for Infinite Hierarchical Change-Point Detection | Change-point detection (CPD) aims to locate abrupt transitions in the
generative model of a sequence of observations. When Bayesian methods are
considered, the standard practice is to infer the posterior distribution of the
change-point locations. However, for complex models (high-dimensional or
heterogeneous), it is not possible to perform reliable detection. To circumvent
this problem, we propose to use a hierarchical model, which yields observations
that belong to a lower-dimensional manifold. Concretely, we consider a
latent-class model with an unbounded number of categories, which is based on
the chinese-restaurant process (CRP). For this model we derive a continual
learning mechanism that is based on the sequential construction of the CRP and
the expectation-maximization (EM) algorithm with a stochastic maximization
step. Our results show that the proposed method is able to recursively infer
the number of underlying latent classes and perform CPD in a reliable manner.
| stat |
Mixtures of multivariate generalized linear models with overlapping
clusters | With the advent of ubiquitous monitoring and measurement protocols, studies
have started to focus more and more on complex, multivariate and heterogeneous
datasets. In such studies, multivariate response variables are drawn from a
heterogeneous population often in the presence of additional covariate
information. In order to deal with this intrinsic heterogeneity, regression
analyses have to be clustered for different groups of units. Up until now,
mixture model approaches assigned units to distinct and non-overlapping groups.
However, not rarely these units exhibit more complex organization and
clustering. It is our aim to define a mixture of generalized linear models with
overlapping clusters of units. This involves crucially an overlap function,
that maps the coefficients of the parent clusters into the the coefficient of
the multiple allocation units. We present a computationally efficient MCMC
scheme that samples the posterior distribution of the parameters in the model.
An example on a two-mode network study shows details of the implementation in
the case of a multivariate probit regression setting. A simulation study shows
the overall performance of the method, whereas an illustration of the voting
behaviour on the US supreme court shows how the 9 justices split in two
overlapping sets of justices.
| stat |
Challenging common bolus advisor for self-monitoring type-I diabetes
patients using Reinforcement Learning | Patients with diabetes who are self-monitoring have to decide right before
each meal how much insulin they should take. A standard bolus advisor exists,
but has never actually been proven to be optimal in any sense. We challenged
this rule applying Reinforcement Learning techniques on data simulated with
T1DM, an FDA-approved simulator developed by Kovatchev et al. modeling the
gluco-insulin interaction. Results show that the optimal bolus rule is fairly
different from the standard bolus advisor, and if followed can actually avoid
hypoglycemia episodes.
| stat |
Probabilistic Forecasting of Temporal Trajectories of Regional Power
Production - Part 2: Photovoltaic Solar | We propose a fully probabilistic prediction model for spatially aggregated
solar photovoltaic (PV) power production at an hourly time scale with lead
times up to several days using weather forecasts from numerical weather
prediction systems as covariates. After an appropriate logarithmic
transformation of the power production, we develop a multivariate Gaussian
prediction model under a Bayesian inference framework. The model incorporates
the temporal error correlation yielding physically consistent forecast
trajectories. Several formulations of the correlation structure are proposed
and investigated. Our method is one of a few approaches that issue full
predictive distributions for PV power production. In a case study of PV power
production in Germany, the method gives calibrated and skillful forecasts.
| stat |
A standardized framework for risk-based assessment of treatment effect
heterogeneity in observational healthcare databases | Aim: One of the aims of the Observation Health Data Sciences and Informatics
(OHDSI) initiative is population-level treatment effect estimation in large
observational databases. Since treatment effects are well-known to vary across
groups of patients with different baseline risk, we aimed to extend the OHDSI
methods library with a framework for risk-based assessment of treatment effect
heterogeneity.
Materials and Methods: The proposed framework consists of five steps: 1)
definition of the problem, i.e. the population, the treatment, the comparator
and the outcome(s) of interest; 2) identification of relevant databases; 3)
development of a prediction model for the outcome(s) of interest; 4) estimation
of propensity scores within strata of predicted risk and estimation of relative
and absolute treatment effect within strata of predicted risk; 5) evaluation
and presentation of results.
Results: We demonstrate our framework by evaluating heterogeneity of the
effect of angiotensin-converting enzyme (ACE) inhibitors versus beta blockers
on a set of 9 outcomes of interest across three observational databases. With
increasing risk of acute myocardial infarction we observed increasing absolute
benefits, i.e. from -0.03% to 0.54% in the lowest to highest risk groups.
Cough-related absolute harms decreased from 4.1% to 2.6%.
Conclusions: The proposed framework may be useful for the evaluation of
heterogeneity of treatment effect on observational data that are mapped to the
OMOP Common Data Model. The proof of concept study demonstrates its feasibility
in large observational data. Further insights may arise by application to
safety and effectiveness questions across the global data network.
| stat |
When deep learning meets causal inference: a computational framework for
drug repurposing from real-world data | Drug repurposing is an effective strategy to identify new uses for existing
drugs, providing the quickest possible transition from bench to bedside.
Existing methods for drug repurposing that mainly focus on pre-clinical
information may exist translational issues when applied to human beings. Real
world data (RWD), such as electronic health records and insurance claims,
provide information on large cohorts of users for many drugs. Here we present
an efficient and easily-customized framework for generating and testing
multiple candidates for drug repurposing using a retrospective analysis of
RWDs. Building upon well-established causal inference and deep learning
methods, our framework emulates randomized clinical trials for drugs present in
a large-scale medical claims database. We demonstrate our framework in a case
study of coronary artery disease (CAD) by evaluating the effect of 55
repurposing drug candidates on various disease outcomes. We achieve 6 drug
candidates that significantly improve the CAD outcomes but not have been
indicated for treating CAD, paving the way for drug repurposing.
| stat |
Random Geometric Graphs on Euclidean Balls | We consider a latent space model for random graphs where a node $i$ is
associated to a random latent point $X_i$ on the Euclidean unit ball. The
probability that an edge exists between two nodes is determined by a ``link''
function, which corresponds to a dot product kernel. For a given class $\F$ of
spherically symmetric distributions for $X_i$, we consider two estimation
problems: latent norm recovery and latent Gram matrix estimation. We construct
an estimator for the latent norms based on the degree of the nodes of an
observed graph in the case of the model where the edge probability is given by
$f(\langle X_i,X_j\rangle)=\mathbbm{1}_{\langle X_i,X_j\rangle\geq \tau}$,
where $0<\tau<1$. We introduce an estimator for the Gram matrix based on the
eigenvectors of observed graph and we establish Frobenius type guarantee for
the error, provided that the link function is sufficiently regular in the
Sobolev sense and that a spectral-gap-type condition holds. We prove that for
certain link functions, the model considered here generates graphs with degree
distribution that have tails with a power-law-type distribution, which can be
seen as an advantage of the model presented here with respect to the classic
Random Geometric Graph model on the Euclidean sphere. We illustrate our results
with numerical experiments.
| stat |
Dynamic mode decomposition in vector-valued reproducing kernel Hilbert
spaces for extracting dynamical structure among observables | Understanding nonlinear dynamical systems (NLDSs) is challenging in a variety
of engineering and scientific fields. Dynamic mode decomposition (DMD), which
is a numerical algorithm for the spectral analysis of Koopman operators, has
been attracting attention as a way of obtaining global modal descriptions of
NLDSs without requiring explicit prior knowledge. However, since existing DMD
algorithms are in principle formulated based on the concatenation of scalar
observables, it is not directly applicable to data with dependent structures
among observables, which take, for example, the form of a sequence of graphs.
In this paper, we formulate Koopman spectral analysis for NLDSs with structures
among observables and propose an estimation algorithm for this problem. This
method can extract and visualize the underlying low-dimensional global dynamics
of NLDSs with structures among observables from data, which can be useful in
understanding the underlying dynamics of such NLDSs. To this end, we first
formulate the problem of estimating spectra of the Koopman operator defined in
vector-valued reproducing kernel Hilbert spaces, and then develop an estimation
procedure for this problem by reformulating tensor-based DMD. As a special case
of our method, we propose the method named as Graph DMD, which is a numerical
algorithm for Koopman spectral analysis of graph dynamical systems, using a
sequence of adjacency matrices. We investigate the empirical performance of our
method by using synthetic and real-world data.
| stat |
LASSO risk and phase transition under dependence | We consider the problem of recovering a $k$-sparse signal
${\mbox{$\beta$}}_0\in\mathbb{R}^p$ from noisy observations $\bf y={\bf
X}\mbox{$\beta$}_0+{\bf w}\in\mathbb{R}^n$. One of the most popular approaches
is the $l_1$-regularized least squares, also known as LASSO. We analyze the
mean square error of LASSO in the case of random designs in which each row of
${\bf X}$ is drawn from distribution $N(0,{\mbox{$\Sigma$}})$ with general
${\mbox{$\Sigma$}}$. We first derive the asymptotic risk of LASSO in the limit
of $n,p\rightarrow\infty$ with $n/p\rightarrow\delta$. We then examine
conditions on $n$, $p$, and $k$ for LASSO to exactly reconstruct
${\mbox{$\beta$}}_0$ in the noiseless case ${\bf w}=0$. A phase boundary
$\delta_c=\delta(\epsilon)$ is precisely established in the phase space defined
by $0\le\delta,\epsilon\le 1$, where $\epsilon=k/p$. Above this boundary, LASSO
perfectly recovers ${\mbox{$\beta$}}_0$ with high probability. Below this
boundary, LASSO fails to recover $\mbox{$\beta$}_0$ with high probability.
While the values of the non-zero elements of ${\mbox{$\beta$}}_0$ do not have
any effect on the phase transition curve, our analysis shows that $\delta_c$
does depend on the signed pattern of the nonzero values of $\mbox{$\beta$}_0$
for general ${\mbox{$\Sigma$}}\ne{\bf I}_p$. This is in sharp contrast to the
previous phase transition results derived in i.i.d. case with
$\mbox{$\Sigma$}={\bf I}_p$ where $\delta_c$ is completely determined by
$\epsilon$ regardless of the distribution of $\mbox{$\beta$}_0$. Underlying our
formalism is a recently developed efficient algorithm called approximate
message passing (AMP) algorithm. We generalize the state evolution of AMP from
i.i.d. case to general case with ${\mbox{$\Sigma$}}\ne{\bf I}_p$. Extensive
computational experiments confirm that our theoretical predictions are
consistent with simulation results on moderate size system.
| stat |
Co-modularity and Detection of Co-communities | This paper introduces the notion of co-modularity, to co-cluster observations
of bipartite networks into co-communities. The task of co-clustering is to
group together nodes of one type, whose interaction with nodes of another type
are the most similar. The novel measure of co-modularity is introduced to
assess the strength of co-communities, as well as to arrange the representation
of nodes and clusters for visualisation, and to define an objective function
for optimisation. We illustrate the power of our proposed methodology on
simulated data, as well as with examples from genomics and consumer-product
reviews.
| stat |
Applications of band-limited extrapolation to forecasting of weather and
financial time series | This paper describes the practical application of causal extrapolation of
sequences for the purpose of forecasting. The methods and proofs have been
applied to simulations to measure the range which data can be accurately
extrapolated. Real world data from the Australian Stock exchange and the
Australian Bureau of Meteorology have been tested and compared with simple
linear extrapolation of the same data. In a majority of the tested scenarios
casual extrapolation has been proved to be the more effective forecaster.
| stat |
No-PASt-BO: Normalized Portfolio Allocation Strategy for Bayesian
Optimization | Bayesian Optimization (BO) is a framework for black-box optimization that is
especially suitable for expensive cost functions. Among the main parts of a BO
algorithm, the acquisition function is of fundamental importance, since it
guides the optimization algorithm by translating the uncertainty of the
regression model in a utility measure for each point to be evaluated.
Considering such aspect, selection and design of acquisition functions are one
of the most popular research topics in BO. Since no single acquisition function
was proved to have better performance in all tasks, a well-established approach
consists of selecting different acquisition functions along the iterations of a
BO execution. In such an approach, the GP-Hedge algorithm is a widely used
option given its simplicity and good performance. Despite its success in
various applications, GP-Hedge shows an undesirable characteristic of
accounting on all past performance measures of each acquisition function to
select the next function to be used. In this case, good or bad values obtained
in an initial iteration may impact the choice of the acquisition function for
the rest of the algorithm. This fact may induce a dominant behavior of an
acquisition function and impact the final performance of the method. Aiming to
overcome such limitation, in this work we propose a variant of GP-Hedge, named
No-PASt-BO, that reduce the influence of far past evaluations. Moreover, our
method presents a built-in normalization that avoids the functions in the
portfolio to have similar probabilities, thus improving the exploration. The
obtained results on both synthetic and real-world optimization tasks indicate
that No-PASt-BO presents competitive performance and always outperforms
GP-Hedge.
| stat |
Identification of COVID-19 mortality patterns in Brazil by a functional
QR decomposition analysis | This paper introduces a functional extension of the QR decomposition of
linear algebra and shows its application to identify independent patterns
(curves) of COVID-19 mortality in Brazil's states. The problem is treated as a
subset selection one, and regions of influence of each pattern are then
determined by fitting the mortality curves of the remaining states to the main
independent ones. Three main patterns are detected: (1) a two-peak curve in
central and southern states, (2) a curve with an early single peak concentrated
in the Amazonian state of Roraima, and (3) a curve with and early peak and a
large recent increase in the Amazonian, northeastern and southeastern states.
| stat |
Modeling and Computation of High Efficiency and Efficacy Multi-Step
Batch Testing for Infectious Diseases | We propose a mathematical model based on probability theory to optimize
COVID-19 testing by a multi-step batch testing approach with variable batch
sizes. This model and simulation tool dramatically increase the efficiency and
efficacy of the tests in a large population at a low cost, particularly when
the infection rate is low. The proposed method combines statistical modeling
with numerical methods to solve nonlinear equations and obtain optimal batch
sizes at each step of tests, with the flexibility to incorporate geographic and
demographic information. In theory, this method substantially improves the
false positive rate and positive predictive value as well. We also conducted a
Monte Carlo simulation to verify this theory. Our simulation results show that
our method significantly reduces the false negative rate. More accurate
assessment can be made if the dilution effect or other practical factors are
taken into consideration. The proposed method will be particularly useful for
the early detection of infectious diseases and prevention of future pandemics.
The proposed work will have broader impacts on medical testing for contagious
diseases in general.
| stat |
The correlation-assisted missing data estimator | We introduce a novel approach to estimation problems in settings with missing
data. Our proposal -- the Correlation-Assisted Missing data (CAM) estimator --
works by exploiting the relationship between the observations with missing
features and those without missing features in order to obtain improved
prediction accuracy. In particular, our theoretical results elucidate general
conditions under which the proposed CAM estimator has lower mean squared error
than the widely used complete-case approach in a range of estimation problems.
We showcase in detail how the CAM estimator can be applied to $U$-Statistics to
obtain an unbiased, asymptotically Gaussian estimator that has lower variance
than the complete-case $U$-Statistic. Further, in nonparametric density
estimation and regression problems, we construct our CAM estimator using kernel
functions, and show it has lower asymptotic mean squared error than the
corresponding complete-case kernel estimator. We also include practical
demonstrations throughout the paper using simulated data and the Terneuzen
birth cohort and Brandsma datasets available from CRAN.
| stat |
Noise-Induced Randomization in Regression Discontinuity Designs | Regression discontinuity designs are used to estimate causal effects in
settings where treatment is determined by whether an observed running variable
crosses a pre-specified threshold. While the resulting sampling design is
sometimes described as akin to a locally randomized experiment in a
neighborhood of the threshold, standard formal analyses do not make reference
to probabilistic treatment assignment and instead identify treatment effects
via continuity arguments. Here we propose a new approach to identification,
estimation, and inference in regression discontinuity designs that exploits
measurement error in the running variable. Under an assumption that the
measurement error is exogenous, we show how to consistently estimate causal
effects using a class of linear estimators that weight treated and control
units so as to balance a latent variable of which the running variable is a
noisy measure. We find this approach to facilitate identification of both
familiar estimands from the literature, as well as policy-relevant estimands
that correspond to the effects of realistic changes to the existing treatment
assignment rule. We demonstrate the method with a study of retention of HIV
patients and evaluate its performance using simulated data and a regression
discontinuity design artificially constructed from test scores in early
childhood.
| stat |
Tensor Factor Model Estimation by Iterative Projection | Tensor time series, which is a time series consisting of tensorial
observations, has become ubiquitous. It typically exhibits high dimensionality.
One approach for dimension reduction is to use a factor model structure, in a
form similar to Tucker tensor decomposition, except that the time dimension is
treated as a dynamic process with a time dependent structure. In this paper we
introduce two approaches to estimate such a tensor factor model by using
iterative orthogonal projections of the original tensor time series. The
approaches extend the existing estimation procedures and our theoretical
investigation shows that they improve the estimation accuracy and convergence
rate significantly. The developed approaches are similar to higher order
orthogonal projection methods for tensor decomposition, but with significant
differences and theoretical properties. Simulation study is conducted to
further illustrate the statistical properties of these estimators.
| stat |
Sufficient Representations for Categorical Variables | Many learning algorithms require categorical data to be transformed into real
vectors before it can be used as input. Often, categorical variables are
encoded as one-hot (or dummy) vectors. However, this mode of representation can
be wasteful since it adds many low-signal regressors, especially when the
number of unique categories is large. In this paper, we investigate simple
alternative solutions for universally consistent estimators that rely on
lower-dimensional real-valued representations of categorical variables that are
"sufficient" in the sense that no predictive information is lost. We then
compare preexisting and proposed methods on simulated and observational
datasets.
| stat |
Approximate Variational Inference Based on a Finite Sample of Gaussian
Latent Variables | Variational methods are employed in situations where exact Bayesian inference
becomes intractable due to the difficulty in performing certain integrals.
Typically, variational methods postulate a tractable posterior and formulate a
lower bound on the desired integral to be approximated, e.g. marginal
likelihood. The lower bound is then optimised with respect to its free
parameters, the so called variational parameters. However, this is not always
possible as for certain integrals it is very challenging (or tedious) to come
up with a suitable lower bound. Here we propose a simple scheme that overcomes
some of the awkward cases where the usual variational treatment becomes
difficult. The scheme relies on a rewriting of the lower bound on the model
log-likelihood. We demonstrate the proposed scheme on a number of synthetic and
real examples, as well as on a real geophysical model for which the standard
variational approaches are inapplicable.
| stat |
Differential Network Analysis: A Statistical Perspective | Networks effectively capture interactions among components of complex
systems, and have thus become a mainstay in many scientific disciplines.
Growing evidence, especially from biology, suggest that networks undergo
changes over time, and in response to external stimuli. In biology and
medicine, these changes have been found to be predictive of complex diseases.
They have also been used to gain insight into mechanisms of disease initiation
and progression. Primarily motivated by biological applications, this article
provides a review of recent statistical machine learning methods for inferring
networks and identifying changes in their structures.
| stat |
Data Driven Reaction Mechanism Estimation via Transient Kinetics and
Machine Learning | Understanding the set of elementary steps and kinetics in each reaction is
extremely valuable to make informed decisions about creating the next
generation of catalytic materials. With physical and mechanistic complexity of
industrial catalysts, it is critical to obtain kinetic information through
experimental methods. As such, this work details a methodology based on the
combination of transient rate/concentration dependencies and machine learning
to measure the number of active sites, the individual rate constants, and gain
insight into the mechanism under a complex set of elementary steps. This new
methodology was applied to simulated transient responses to verify its ability
to obtain correct estimates of the micro-kinetic coefficients. Furthermore,
experimental CO oxidation data was analyzed to reveal the Langmuir-Hinshelwood
mechanism driving the reaction. As oxygen accumulated on the catalyst, a
transition in the mechanism was clearly defined in the machine learning
analysis due to the large amount of kinetic information available from
transient reaction techniques. This methodology is proposed as a new data
driven approach to characterize how materials control complex reaction
mechanisms relying exclusively on experimental data.
| stat |
Multiple imputation using chained random forests: a preliminary study
based on the empirical distribution of out-of-bag prediction errors | Missing data are common in data analyses in biomedical fields, and imputation
methods based on random forests (RF) have become widely accepted, as the RF
algorithm can achieve high accuracy without the need for specification of data
distributions or relationships. However, the predictions from RF do not contain
information about prediction uncertainty, which was unacceptable for multiple
imputation. Available RF-based multiple imputation methods tried to do proper
multiple imputation either by sampling directly from observations under
predicting nodes without accounting for the prediction error or by making
normality assumption about the prediction error distribution. In this study, a
novel RF-based multiple imputation method was proposed by constructing
conditional distributions the empirical distribution of out-of-bag prediction
errors. The proposed method was compared with previous method with parametric
assumptions about RF's prediction errors and predictive mean matching based on
simulation studies on data with presence of interaction term. The proposed
non-parametric method can deliver valid multiple imputation results. The
accompanying R package for this study is publicly available.
| stat |
Statistical Inference for Online Decision Making via Stochastic Gradient
Descent | Online decision making aims to learn the optimal decision rule by making
personalized decisions and updating the decision rule recursively. It has
become easier than before with the help of big data, but new challenges also
come along. Since the decision rule should be updated once per step, an offline
update which uses all the historical data is inefficient in computation and
storage. To this end, we propose a completely online algorithm that can make
decisions and update the decision rule online via stochastic gradient descent.
It is not only efficient but also supports all kinds of parametric reward
models. Focusing on the statistical inference of online decision making, we
establish the asymptotic normality of the parameter estimator produced by our
algorithm and the online inverse probability weighted value estimator we used
to estimate the optimal value. Online plugin estimators for the variance of the
parameter and value estimators are also provided and shown to be consistent, so
that interval estimation and hypothesis test are possible using our method. The
proposed algorithm and theoretical results are tested by simulations and a real
data application to news article recommendation.
| stat |
An Expectation Maximization Framework for Yule-Simon Preferential
Attachment Models | In this paper we develop an Expectation Maximization(EM) algorithm to
estimate the parameter of a Yule-Simon distribution. The Yule-Simon
distribution exhibits the "rich get richer" effect whereby an 80-20 type of
rule tends to dominate. These distributions are ubiquitous in industrial
settings. The EM algorithm presented provides both frequentist and Bayesian
estimates of the $\lambda$ parameter. By placing the estimation method within
the EM framework we are able to derive Standard errors of the resulting
estimate. Additionally, we prove convergence of the Yule-Simon EM algorithm and
study the rate of convergence. An explicit, closed form solution for the rate
of convergence of the algorithm is given. Applications including graph node
degree distribution estimation are listed.
| stat |
Objective Bayesian meta-analysis based on generalized multivariate
random effects model | Objective Bayesian inference procedures are derived for the parameters of the
multivariate random effects model generalized to elliptically contoured
distributions. The posterior for the overall mean vector and the between-study
covariance matrix is deduced by assigning two noninformative priors to the
model parameter, namely the Berger and Bernardo reference prior and the
Jeffreys prior, whose analytical expressions are obtained under weak
distributional assumptions. It is shown that the only condition needed for the
posterior to be proper is that the sample size is larger than the dimension of
the data-generating model, independently of the class of elliptically contoured
distributions used in the definition of the generalized multivariate random
effects model. The theoretical findings of the paper are applied to real data
consisting of ten studies about the effectiveness of hypertension treatment for
reducing blood pressure where the treatment effects on both the systolic blood
pressure and diastolic blood pressure are investigated.
| stat |
Uncertainty quantification for nonconvex tensor completion: Confidence
intervals, heteroscedasticity and optimality | We study the distribution and uncertainty of nonconvex optimization for noisy
tensor completion -- the problem of estimating a low-rank tensor given
incomplete and corrupted observations of its entries. Focusing on a two-stage
estimation algorithm proposed by Cai et al. (2019), we characterize the
distribution of this nonconvex estimator down to fine scales. This
distributional theory in turn allows one to construct valid and short
confidence intervals for both the unseen tensor entries and the unknown tensor
factors. The proposed inferential procedure enjoys several important features:
(1) it is fully adaptive to noise heteroscedasticity, and (2) it is data-driven
and automatically adapts to unknown noise distributions. Furthermore, our
findings unveil the statistical optimality of nonconvex tensor completion: it
attains un-improvable $\ell_{2}$ accuracy -- including both the rates and the
pre-constants -- when estimating both the unknown tensor and the underlying
tensor factors.
| stat |
Ensemble learning reveals dissimilarity between rare-earth transition
metal binary alloys with respect to the Curie temperature | We propose a data-driven method to extract dissimilarity between materials,
with respect to a given target physical property. The technique is based on an
ensemble method with Kernel ridge regression as the predicting model; multiple
random subset sampling of the materials is done to generate prediction models
and the corresponding contributions of the reference training materials in
detail. The distribution of the predicted values for each material can be
approximated by a Gaussian mixture model. The reference training materials
contributed to the prediction model that accurately predicts the physical
property value of a specific material, are considered to be similar to that
material, or vice versa. Evaluations using synthesized data demonstrate that
the proposed method can effectively measure the dissimilarity between data
instances. An application of the analysis method on the data of Curie
temperature (TC) of binary 3d transition metal 4f rare earth binary alloys also
reveals meaningful results on the relations between the materials. The proposed
method can be considered as a potential tool for obtaining a deeper
understanding of the structure of data, with respect to a target property, in
particular.
| stat |
Attraction-Repulsion clustering with applications to fairness | In the framework of fair learning, we consider clustering methods that avoid
or limit the influence of a set of protected attributes, $S$, (race, sex, etc)
over the resulting clusters, with the goal of producing a {\it fair
clustering}. For this, we introduce perturbations to the Euclidean distance
that take into account $S$ in a way that resembles attraction-repulsion in
charged particles in Physics and results in dissimilarities with an easy
interpretation. Cluster analysis based on these dissimilarities penalizes
homogeneity of the clusters in the attributes $S$, and leads to an improvement
in fairness. We illustrate the use of our procedures with both synthetic and
real data. Our procedures are implemented in an R package freely available at
https://github.com/HristoInouzhe/AttractionRepulsionClustering.
| stat |
Simultaneous Inference for Multiple Proportions: A Multivariate
Beta-Binomial Model | Statistical inference in high-dimensional settings is challenging when
standard unregularized methods are employed. In this work, we focus on the case
of multiple correlated proportions for which we develop a Bayesian inference
framework. For this purpose, we construct an $m$-dimensional Beta distribution
from a $2^m$-dimensional Dirichlet distribution, building on work by Olkin and
Trikalinos (2015). This readily leads to a multivariate Beta-binomial model for
which simple update rules from the common Dirichlet-multinomial model can be
adopted. From the frequentist perspective, this approach amounts to adding
pseudo-observations to the data and allows a joint shrinkage estimation of mean
vector and covariance matrix. For higher dimensions ($m > 10$), the extensive
model based on $2^m$ parameters starts to become numerically infeasible. To
counter this problem, we utilize a reduced parametrisation which has only $1 +
m(m + 1)/2$ parameters describing first and second order moments. A copula
model can then be used to approximate the (posterior) multivariate Beta
distribution. A natural inference goal is the construction of multivariate
credible regions. The properties of different credible regions are assessed in
a simulation study in the context of investigating the accuracy of multiple
binary classifiers. It is shown that the extensive and copula approach lead to
a (Bayes) coverage probability very close to the target level. In this regard,
they outperform credible regions based on a normal approximation of the
posterior distribution, in particular for small sample sizes. Additionally,
they always lead to credible regions which lie entirely in the parameter space
which is not the case when the normal approximation is used.
| stat |
Shannon's entropy and its Generalizations towards Statistics,
Reliability and Information Science during 1948-2018 | Starting from the pioneering works of Shannon and Weiner in 1948, a plethora
of works have been reported on entropy in different directions. Entropy-related
review work in the direction of statistics, reliability and information
science, to the best of our knowledge, has not been reported so far. Here we
have tried to collect all possible works in this direction during the period
1948-2018 so that people interested in entropy, specially the new researchers,
get benefited.
| stat |
Sequential Neural Likelihood: Fast Likelihood-free Inference with
Autoregressive Flows | We present Sequential Neural Likelihood (SNL), a new method for Bayesian
inference in simulator models, where the likelihood is intractable but
simulating data from the model is possible. SNL trains an autoregressive flow
on simulated data in order to learn a model of the likelihood in the region of
high posterior density. A sequential training procedure guides simulations and
reduces simulation cost by orders of magnitude. We show that SNL is more
robust, more accurate and requires less tuning than related neural-based
methods, and we discuss diagnostics for assessing calibration, convergence and
goodness-of-fit.
| stat |
Bayesian Optimization using Pseudo-Points | Bayesian optimization (BO) is a popular approach for expensive black-box
optimization, with applications including parameter tuning, experimental
design, robotics. BO usually models the objective function by a Gaussian
process (GP), and iteratively samples the next data point by maximizing an
acquisition function. In this paper, we propose a new general framework for BO
by generating pseudo-points (i.e., data points whose objective values are not
evaluated) to improve the GP model. With the classic acquisition function,
i.e., upper confidence bound (UCB), we prove that the cumulative regret can be
generally upper bounded. Experiments using UCB and other acquisition functions,
i.e., probability of improvement (PI) and expectation of improvement (EI), on
synthetic as well as real-world problems clearly show the advantage of
generating pseudo-points.
| stat |
Detecting conflicting summary statistics in likelihood-free inference | Bayesian likelihood-free methods implement Bayesian inference using
simulation of data from the model to substitute for intractable likelihood
evaluations. Most likelihood-free inference methods replace the full data set
with a summary statistic before performing Bayesian inference, and the choice
of this statistic is often difficult. The summary statistic should be
low-dimensional for computational reasons, while retaining as much information
as possible about the parameter. Using a recent idea from the interpretable
machine learning literature, we develop some regression-based diagnostic
methods which are useful for detecting when different parts of a summary
statistic vector contain conflicting information about the model parameters.
Conflicts of this kind complicate summary statistic choice, and detecting them
can be insightful about model deficiencies and guide model improvement. The
diagnostic methods developed are based on regression approaches to
likelihood-free inference, in which the regression model estimates the
posterior density using summary statistics as features. Deletion and imputation
of part of the summary statistic vector within the regression model can remove
conflicts and approximate posterior distributions for summary statistic
subsets. A larger than expected change in the estimated posterior density
following deletion and imputation can indicate a conflict in which inferences
of interest are affected. The usefulness of the new methods is demonstrated in
a number of real examples.
| stat |
Winning Is Not Everything: A contextual analysis of hockey face-offs | This paper takes a different approach to evaluating face-offs in ice hockey.
Instead of looking at win percentages, the de facto measure of successful
face-off takers for decades, focuses on the game events following the face-off
and how directionality, clean wins, and player handedness play a significant
role in creating value. This will demonstrate how not all face-off wins are
made equal: some players consistently create post-face-off value through clean
wins and by directing the puck to high-value areas of the ice. As a result, we
propose an expected events face-off model as well as a wins above expected
model that take into account the value added on a face-off by targeting the
puck to specific areas on the ice in various contexts, as well as the impact
this has on subsequent game events.
| stat |
Evaluating the Disentanglement of Deep Generative Models through
Manifold Topology | Learning disentangled representations is regarded as a fundamental task for
improving the generalization, robustness, and interpretability of generative
models. However, measuring disentanglement has been challenging and
inconsistent, often dependent on an ad-hoc external model or specific to a
certain dataset. To address this, we present a method for quantifying
disentanglement that only uses the generative model, by measuring the
topological similarity of conditional submanifolds in the learned
representation. This method showcases both unsupervised and supervised
variants. To illustrate the effectiveness and applicability of our method, we
empirically evaluate several state-of-the-art models across multiple datasets.
We find that our method ranks models similarly to existing methods. We make
ourcode publicly available at
https://github.com/stanfordmlgroup/disentanglement.
| stat |
On the Convergence of Extended Variational Inference for Non-Gaussian
Statistical Models | Variational inference (VI) is a widely used framework in Bayesian estimation.
For most of the non-Gaussian statistical models, it is infeasible to find an
analytically tractable solution to estimate the posterior distributions of the
parameters. Recently, an improved framework, namely the extended variational
inference (EVI), has been introduced and applied to derive analytically
tractable solution by employing lower-bound approximation to the variational
objective function. Two conditions required for EVI implementation, namely the
weak condition and the strong condition, are discussed and compared in this
paper. In practical implementation, the convergence of the EVI depends on the
selection of the lower-bound approximation, no matter with the weak condition
or the strong condition. In general, two approximation strategies, the single
lower-bound (SLB) approximation and the multiple lower-bounds (MLB)
approximation, can be applied to carry out the lower-bound approximation. To
clarify the differences between the SLB and the MLB, we will also discuss the
convergence properties of the aforementioned two approximations. Extensive
comparisons are made based on some existing EVI-based non-Gaussian statistical
models. Theoretical analysis are conducted to demonstrate the differences
between the weak and the strong conditions. Qualitative and quantitative
experimental results are presented to show the advantages of the SLB
approximation.
| stat |
Tuning-free ridge estimators for high-dimensional generalized linear
models | Ridge estimators regularize the squared Euclidean lengths of parameters. Such
estimators are mathematically and computationally attractive but involve tuning
parameters that can be difficult to calibrate. In this paper, we show that
ridge estimators can be modified such that tuning parameters can be avoided
altogether. We also show that these modified versions can improve on the
empirical prediction accuracies of standard ridge estimators combined with
cross-validation, and we provide first theoretical guarantees.
| stat |
Inference and Uncertainty Quantification for Noisy Matrix Completion | Noisy matrix completion aims at estimating a low-rank matrix given only
partial and corrupted entries. Despite substantial progress in designing
efficient estimation algorithms, it remains largely unclear how to assess the
uncertainty of the obtained estimates and how to perform statistical inference
on the unknown matrix (e.g.~constructing a valid and short confidence interval
for an unseen entry).
This paper takes a step towards inference and uncertainty quantification for
noisy matrix completion. We develop a simple procedure to compensate for the
bias of the widely used convex and nonconvex estimators. The resulting
de-biased estimators admit nearly precise non-asymptotic distributional
characterizations, which in turn enable optimal construction of confidence
intervals\,/\,regions for, say, the missing entries and the low-rank factors.
Our inferential procedures do not rely on sample splitting, thus avoiding
unnecessary loss of data efficiency. As a byproduct, we obtain a sharp
characterization of the estimation accuracy of our de-biased estimators, which,
to the best of our knowledge, are the first tractable algorithms that provably
achieve full statistical efficiency (including the preconstant). The analysis
herein is built upon the intimate link between convex and nonconvex
optimization --- an appealing feature recently discovered by
\cite{chen2019noisy}.
| stat |
Analysis of Prescription Drug Utilization with Beta Regression Models | The healthcare sector in the U.S. is complex and is also a large sector that
generates about 20% of the country's gross domestic product. Healthcare
analytics has been used by researchers and practitioners to better understand
the industry. In this paper, we examine and demonstrate the use of Beta
regression models to study the utilization of brand name drugs in the U.S. to
understand the variability of brand name drug utilization across different
areas. The models are fitted to public datasets obtained from the Medicare &
Medicaid Services and the Internal Revenue Service. Integrated Nested Laplace
Approximation (INLA) is used to perform the inference. The numerical results
show that Beta regression models can fit the brand name drug claim rates well
and including spatial dependence improves the performance of the Beta
regression models. Such models can be used to reflect the effect of
prescription drug utilization when updating an insured's health risk in a risk
scoring model.
| stat |
Additive Bayesian variable selection under censoring and
misspecification | We discuss the role of misspecification and censoring on Bayesian model
selection in the contexts of right-censored survival and concave log-likelihood
regression. Misspecification includes wrongly assuming the censoring mechanism
to be non-informative. Emphasis is placed on additive accelerated failure time,
Cox proportional hazards and probit models. We offer a theoretical treatment
that includes local and non-local priors, and a general non-linear effect
decomposition to improve power-sparsity trade-offs. We discuss a fundamental
question: what solution can one hope to obtain when (inevitably) models are
misspecified, and how to interpret it? Asymptotically, covariates that do not
have predictive power for neither the outcome nor (for survival data) censoring
times, in the sense of reducing a likelihood-associated loss, are discarded.
Misspecification and censoring have an asymptotically negligible effect on
false positives, but their impact on power is exponential. We show that it can
be advantageous to consider simple models that are computationally practical
yet attain good power to detect potentially complex effects, including the use
of finite-dimensional basis to detect truly non-parametric effects. We also
discuss algorithms to capitalize on sufficient statistics and fast likelihood
approximations for Gaussian-based survival and binary models.
| stat |
Predicting into unknown space? Estimating the area of applicability of
spatial prediction models | Predictive modelling using machine learning has become very popular for
spatial mapping of the environment. Models are often applied to make
predictions far beyond sampling locations where new geographic locations might
considerably differ from the training data in their environmental properties.
However, areas in the predictor space without support of training data are
problematic. Since the model has no knowledge about these environments,
predictions have to be considered uncertain.
Estimating the area to which a prediction model can be reliably applied is
required. Here, we suggest a methodology that delineates the "area of
applicability" (AOA) that we define as the area, for which the cross-validation
error of the model applies. We first propose a "dissimilarity index" (DI) that
is based on the minimum distance to the training data in the predictor space,
with predictors being weighted by their respective importance in the model. The
AOA is then derived by applying a threshold based on the DI of the training
data where the DI is calculated with respect to the cross-validation strategy
used for model training. We test for the ideal threshold by using simulated
data and compare the prediction error within the AOA with the cross-validation
error of the model. We illustrate the approach using a simulated case study.
Our simulation study suggests a threshold on DI to define the AOA at the .95
quantile of the DI in the training data. Using this threshold, the prediction
error within the AOA is comparable to the cross-validation RMSE of the model,
while the cross-validation error does not apply outside the AOA. This applies
to models being trained with randomly distributed training data, as well as
when training data are clustered in space and where spatial cross-validation is
applied.
We suggest to report the AOA alongside predictions, complementary to
validation measures.
| stat |
Streamlined Variational Inference for Linear Mixed Models with Crossed
Random Effects | We derive streamlined mean field variational Bayes algorithms for fitting
linear mixed models with crossed random effects. In the most general situation,
where the dimensions of the crossed groups are arbitrarily large, streamlining
is hindered by lack of sparseness in the underlying least squares system.
Because of this fact we also consider a hierarchy of relaxations of the mean
field product restriction. The least stringent product restriction delivers a
high degree of inferential accuracy. However, this accuracy must be mitigated
against its higher storage and computing demands. Faster sparse storage and
computing alternatives are also provided, but come with the price of diminished
inferential accuracy. This article provides full algorithmic details of three
variational inference strategies, presents detailed empirical results on their
pros and cons and, thus, guides the users on their choice of variational
inference approach depending on the problem size and computing resources.
| stat |
On a 'Two Truths' Phenomenon in Spectral Graph Clustering | Clustering is concerned with coherently grouping observations without any
explicit concept of true groupings. Spectral graph clustering - clustering the
vertices of a graph based on their spectral embedding - is commonly approached
via K-means (or, more generally, Gaussian mixture model) clustering composed
with either Laplacian or Adjacency spectral embedding (LSE or ASE). Recent
theoretical results provide new understanding of the problem and solutions, and
lead us to a 'Two Truths' LSE vs. ASE spectral graph clustering phenomenon
convincingly illustrated here via a diffusion MRI connectome data set: the
different embedding methods yield different clustering results, with LSE
capturing left hemisphere/right hemisphere affinity structure and ASE capturing
gray matter/white matter core-periphery structure.
| stat |
Variational Approximation of Factor Stochastic Volatility Models | Estimation and prediction in high dimensional multivariate factor stochastic
volatility models is an important and active research area because such models
allow a parsimonious representation of multivariate stochastic volatility.
Bayesian inference for factor stochastic volatility models is usually done by
Markov chain Monte Carlo methods, often by particle Markov chain Monte Carlo,
which are usually slow for high dimensional or long time series because of the
large number of parameters and latent states involved. Our article makes two
contributions. The first is to propose fast and accurate variational Bayes
methods to approximate the posterior distribution of the states and parameters
in factor stochastic volatility models. The second contribution is to extend
this batch methodology to develop fast sequential variational updates for
prediction as new observations arrive. The methods are applied to simulated and
real datasets and shown to produce good approximate inference and prediction
compared to the latest particle Markov chain Monte Carlo approaches, but are
much faster.
| stat |
Modelling conditional probabilities with Riemann-Theta Boltzmann
Machines | The probability density function for the visible sector of a Riemann-Theta
Boltzmann machine can be taken conditional on a subset of the visible units. We
derive that the corresponding conditional density function is given by a
reparameterization of the Riemann-Theta Boltzmann machine modelling the
original probability density function. Therefore the conditional densities can
be directly inferred from the Riemann-Theta Boltzmann machine.
| stat |
Efficient and Robust Estimation of Linear Regression with Normal Errors | Linear regression with normally distributed errors - including particular
cases such as ANOVA, Student's t-test or location-scale inference - is a widely
used statistical procedure. In this case the ordinary least squares estimator
possesses remarkable properties but is very sensitive to outliers. Several
robust alternatives have been proposed, but there is still significant room for
improvement. This paper thus proposes an original method of estimation that
offers the best efficiency simultaneously in the absence and the presence of
outliers, both for the estimation of the regression coefficients and the scale
parameter. The approach first consists in broadening the normal assumption of
the errors to a mixture of the normal and the filtered-log-Pareto (FLP), an
original distribution designed to represent the outliers. The
expectation-maximization (EM) algorithm is then adapted and we obtain the N-FLP
estimators of the regression coefficients, the scale parameter and the
proportion of outliers, along with probabilities of each observation being an
outlier. The performance of the N-FLP estimators is compared with the best
alternatives in an extensive Monte Carlo simulation. The paper demonstrates
that this method of estimation can also be used for a complete robust
inference, including confidence intervals, hypothesis testing and model
selection.
| stat |
Domain Adaptation for One-Class Classification: Monitoring the Health of
Critical Systems Under Limited Information | The failure of a complex and safety critical industrial asset can have
extremely high consequences. Close monitoring for early detection of abnormal
system conditions is therefore required. Data-driven solutions to this problem
have been limited for two reasons: First, safety critical assets are designed
and maintained to be highly reliable and faults are rare. Fault detection can
thus not be solved with supervised learning. Second, complex industrial systems
usually have long lifetime during which they face very different operating
conditions. In the early life of the system, the collected data is probably not
representative of future operating conditions, making it challenging to train a
robust model.
In this paper, we propose a methodology to monitor the systems in their early
life. To do so, we enhance the training dataset with other units from a fleet,
for which longer observations are available. Since each unit has its own
specificity, we propose to extract features made independent of their origin by
three unsupervised feature alignment techniques. First, using a variational
encoder, we impose a shared probabilistic encoder/decoder for both units.
Second, we introduce a new loss designed to conserve inter-point spacial
relationships between the input and the learned features. Last, we propose to
train in an adversarial manner a discriminator on the origin of the features.
Once aligned, the features are fed to a one-class classifier to monitor the
health of the system. By exploring the different combinations of the proposed
alignment strategies, and by testing them on a real case study, a fleet
composed of 112 power plants operated in different geographical locations and
under very different operating regimes, we demonstrate that this alignment is
necessary and beneficial.
| stat |
BIG sampling | Graph sampling is a statistical approach to study real graphs, which
represent the structure of many technological, social or biological phenomena
of interest. We develop bipartite incident graph sampling (BIGS) as a feasible
representation of graph sampling from arbitrary finite graphs. It provides also
a unified treatment of the existing unconventional sampling methods which were
studied separately in the past, including indirect, network and adaptive
cluster sampling. The sufficient and necessary conditions of feasible BIGS
representation are established, given which one can apply a family of
Hansen-Hurwitz type design-unbiased estimators in addition to the standard
Horvitz-Thompson estimator. The approach increases therefore the potentials of
efficiency gains in graph sampling. A general result regarding the relative
efficiency of the two types of estimators is obtained. Numerical examples are
given to illustrate the versatility of the proposed approach.
| stat |
Solving for multi-class: a survey and synthesis | Many of the best statistical classification algorithms are binary classifiers
that can only distinguish between one of two classes. The number of possible
ways of generalizing binary classification to multi-class increases
exponentially with the number of classes. There is some indication that the
best method will depend on the dataset. Hence, we are particularly interested
in data-driven solution design, whether based on prior considerations or on
empirical examination of the data. Here we demonstrate how a recursive control
language can be used to describe a multitude of different partitioning
strategies in multi-class classification, including those in most common use.
We use it both to manually construct new partitioning configurations as well as
to examine those that have been automatically designed.
Eight different strategies were tested on eight different datasets using a
support vector machine (SVM) as the base binary classifier. Numerical results
suggest that a one-size-fits-all solution consisting of one-versus-one is
appropriate for most datasets. Three datasets showed better accuracy using
different methods. The best solution for the most improved dataset exploited a
property of the data to produce an uncertainty coefficient 36\% higher (0.016
absolute gain) than one-vs.-one. For the same dataset, an adaptive solution
that empirically examined the data was also more accurate than one-vs.-one
while being faster.
| stat |
On genetic correlation estimation with summary statistics from
genome-wide association studies | Genome-wide association studies (GWAS) have been widely used to examine the
association between single nucleotide polymorphisms (SNPs) and complex traits,
where both the sample size n and the number of SNPs p can be very large.
Recently, cross-trait polygenic risk score (PRS) method has gained extremely
popular for assessing genetic correlation of complex traits based on GWAS
summary statistics (e.g., SNP effect size). However, empirical evidence has
shown a common bias phenomenon that even highly significant cross-trait PRS can
only account for a very small amount of genetic variance (R^2 often <1%). The
aim of this paper is to develop a novel and powerful method to address the bias
phenomenon of cross-trait PRS. We theoretically show that the estimated genetic
correlation is asymptotically biased towards zero when complex traits are
highly polygenic/omnigenic. When all p SNPs are used to construct PRS, we show
that the asymptotic bias of PRS estimator is independent of the unknown number
of causal SNPs m. We propose a consistent PRS estimator to correct such
asymptotic bias. We also develop a novel estimator of genetic correlation which
is solely based on two sets of GWAS summary statistics. In addition, we
investigate whether or not SNP screening by GWAS p-values can lead to improved
estimation and show the effect of overlapping samples among GWAS. Our results
may help demystify and tackle the puzzling "missing genetic overlap" phenomenon
of cross-trait PRS for dissecting the genetic similarity of closely related
heritable traits. We illustrate the finite sample performance of our
bias-corrected PRS estimator by using both numerical experiments and the UK
Biobank data, in which we assess the genetic correlation between brain white
matter tracts and neuropsychiatric disorders.
| stat |
INTEREST: INteractive Tool for Exploring REsults from Simulation sTudies | Simulation studies allow us to explore the properties of statistical methods.
They provide a powerful tool with a multiplicity of aims; among others:
evaluating and comparing new or existing statistical methods, assessing
violations of modelling assumptions, helping with the understanding of
statistical concepts, and supporting the design of clinical trials. The
increased availability of powerful computational tools and usable software has
contributed to the rise of simulation studies in the current literature.
However, simulation studies involve increasingly complex designs, making it
difficult to provide all relevant results clearly. Dissemination of results
plays a focal role in simulation studies: it can drive applied analysts to use
methods that have been shown to perform well in their settings, guide
researchers to develop new methods in a promising direction, and provide
insights into less established methods. It is crucial that we can digest
relevant results of simulation studies. Therefore, we developed INTEREST: an
INteractive Tool for Exploring REsults from Simulation sTudies. The tool has
been developed using the Shiny framework in R and is available as a web app or
as a standalone package. It requires uploading a tidy format dataset with the
results of a simulation study in R, Stata, SAS, SPSS, or comma-separated
format. A variety of performance measures are estimated automatically along
with Monte Carlo standard errors; results and performance summaries are
displayed both in tabular and graphical fashion, with a wide variety of
available plots. Consequently, the reader can focus on simulation parameters
and estimands of most interest. In conclusion, INTEREST can facilitate the
investigation of results from simulation studies and supplement the reporting
of results, allowing researchers to share detailed results from their
simulations and readers to explore them freely.
| stat |
Explaining the Decline of Child Mortality in 44 Developing Countries: A
Bayesian Extension of Oaxaca Decomposition Methods | We investigate the decline of infant mortality in 42 low and middle income
countries (LMIC) using detailed micro data from 84 Demographic and Health
Surveys. We estimate infant mortality risk for each infant in our data and
develop a novel extension of Oaxaca decomposition to understand the sources of
these changes. We find that the decline in infant mortality is due to a
declining propensity for parents with given characteristics to experience the
death of an infant rather than due to changes in the distributions of these
characteristics over time. Our results suggest that technical progress and
policy health interventions in the form of public goods are the main drivers of
the the recent decline in infant mortality in LMIC.
| stat |
Use of Emergency Departments by Frail Elderly Patients: Temporal
Patterns and Case Complexity | Emergency department (ED) care for frail elderly patients is associated with
an increased use of resources due to their complex medical needs and frequently
difficult psycho-social situation. To better target their needs with specially
trained staff, it is vital to determine the times during which these particular
patients present to the ED. Recent research was inconclusive regarding this
question and the applied methods were limited to coarse time windows. Moreover,
there is little research on time variation of frail ED patients' case
complexity. This study examines differences in arrival rates for frail vs.
non-frail patients in detail and compares case complexity in frail patients
within vs. outside of regular GP working hours. Arrival times and case
variables (admission rate, ED length of stay [LOS], triage level and
comorbidities) were extracted from the EHR of an ED in an urban German teaching
hospital. We employed Poisson time series regression to determine patterns in
hourly arrival rates over the week. Frail elderly patients presented more
likely to the ED during already high frequented hours, especially at midday and
in the afternoon. Case complexity for frail patients was significantly higher
compared to non-frail patients, but varied marginally in time only with respect
to triage level and ED LOS. The results suggest that frailty-attuned emergency
care should be available in EDs during the busiest hours. Based on EHR data,
hospitals thus can tailor their staff needs.
| stat |
Multinomial logistic model for coinfection diagnosis between arbovirus
and malaria in Kedougou | In tropical regions, populations continue to suffer morbidity and mortality
from malaria and arboviral diseases. In Kedougou (Senegal), these illnesses are
all endemic due to the climate and its geographical position. The
co-circulation of malaria parasites and arboviruses can explain the observation
of coinfected cases. Indeed there is strong resemblance in symptoms between
these diseases making problematic targeted medical care of coinfected cases.
This is due to the fact that the origin of illness is not obviously known. Some
cases could be immunized against one or the other of the pathogens, immunity
typically acquired with factors like age and exposure as usual for endemic
area. Then, coinfection needs to be better diagnosed. Using data collected from
patients in Kedougou region, from 2009 to 2013, we adjusted a multinomial
logistic model and selected relevant variables in explaining coinfection
status. We observed specific sets of variables explaining each of the diseases
exclusively and the coinfection. We tested the independence between arboviral
and malaria infections and derived coinfection probabilities from the model
fitting. In case of a coinfection probability greater than a threshold value to
be calibrated on the data, duration of illness above 3 days and age above 10
years-old are mostly indicative of arboviral disease while body temperature
higher than 40{\textdegree}C and presence of nausea or vomiting symptoms during
the rainy season are mostly indicative of malaria disease.
| stat |
A self-organising eigenspace map for time series clustering | This paper presents a novel time series clustering method, the
self-organising eigenspace map (SOEM), based on a generalisation of the
well-known self-organising feature map (SOFM). The SOEM operates on the
eigenspaces of the embedded covariance structures of time series which are
related directly to modes in those time series. Approximate joint
diagonalisation acts as a pseudo-metric across these spaces allowing us to
generalise the SOFM to a neural network with matrix input. The technique is
empirically validated against three sets of experiments; univariate and
multivariate time series clustering, and application to (clustered)
multi-variate time series forecasting. Results indicate that the technique
performs a valid topologically ordered clustering of the time series. The
clustering is superior in comparison to standard benchmarks when the data is
non-aligned, gives the best clustering stage for when used in forecasting, and
can be used with partial/non-overlapping time series, multivariate clustering
and produces a topological representation of the time series objects.
| stat |
A Stein Goodness of fit Test for Exponential Random Graph Models | We propose and analyse a novel nonparametric goodness of fit testing
procedure for exchangeable exponential random graph models (ERGMs) when a
single network realisation is observed. The test determines how likely it is
that the observation is generated from a target unnormalised ERGM density. Our
test statistics are derived from a kernel Stein discrepancy, a divergence
constructed via Steins method using functions in a reproducing kernel Hilbert
space, combined with a discrete Stein operator for ERGMs. The test is a Monte
Carlo test based on simulated networks from the target ERGM. We show
theoretical properties for the testing procedure for a class of ERGMs.
Simulation studies and real network applications are presented.
| stat |
Functional Correlations in the Pursuit of Performance Assessment of
Classifiers | In statistical classification and machine learning, as well as in social and
other sciences, a number of measures of association have been proposed for
assessing and comparing individual classifiers, raters, as well as their
groups. In this paper, we introduce, justify, and explore several new measures
of association, which we call CO-, ANTI- and COANTI-correlation coefficients,
that we demonstrate to be powerful tools for classifying confusion matrices. We
illustrate the performance of these new coefficients using a number of
examples, from which we also conclude that the coefficients are new objects in
the sense that they differ from those already in the literature.
| stat |
Defending Against Model Stealing Attacks with Adaptive Misinformation | Deep Neural Networks (DNNs) are susceptible to model stealing attacks, which
allows a data-limited adversary with no knowledge of the training dataset to
clone the functionality of a target model, just by using black-box query
access. Such attacks are typically carried out by querying the target model
using inputs that are synthetically generated or sampled from a surrogate
dataset to construct a labeled dataset. The adversary can use this labeled
dataset to train a clone model, which achieves a classification accuracy
comparable to that of the target model. We propose "Adaptive Misinformation" to
defend against such model stealing attacks. We identify that all existing model
stealing attacks invariably query the target model with Out-Of-Distribution
(OOD) inputs. By selectively sending incorrect predictions for OOD queries, our
defense substantially degrades the accuracy of the attacker's clone model (by
up to 40%), while minimally impacting the accuracy (<0.5%) for benign users.
Compared to existing defenses, our defense has a significantly better security
vs accuracy trade-off and incurs minimal computational overhead.
| stat |
Improved Small Area Estimation via Compromise Regression Weights | Shrinkage estimates of small domain parameters typically utilize a
combination of a noisy "direct" estimate that only uses data from a specific
small domain and a more stable regression estimate. When the regression model
is misspecified, estimation performance for the noisier domains can suffer due
to substantial shrinkage towards a poorly estimated regression surface. In this
paper, we introduce a new class of robust, empirically-driven regression
weights that target estimation of the small domain means under potential
misspecification of the global regression model. Our regression weights are a
convex combination of the model-based weights associated with the best linear
unbiased predictor (BLUP) and those associated with the observed best predictor
(OBP). The compromise parameter in this convex combination is found by
minimizing a novel, unbiased estimate of the mean-squared prediction error for
the small domain means, and we label the associated small domain estimates the
"compromise best predictor", or CBP. Using a data-adaptive mixture for the
regression weights enables the CBP to possess the robustness of the OBP while
retaining the main advantages of the EBLUP whenever the regression model is
correct. We demonstrate the use of the CBP in an application estimating gait
speed in older adults.
| stat |
Monte Carlo estimators of first-and total-orders Sobol' indices | This study compares the performances of two sampling-based strategies for the
simultaneous estimation of the first-and total-orders variance-based
sensitivity indices (a.k.a Sobol' indices). The first strategy was introduced
by [8] and is the current approach employed by practitioners. The second one
was only recently introduced by the authors of the present article. They both
rely on different estimators of first-and total-orders Sobol' indices. The
asymp-totic normal variances of the two sets of estimators are established and
their accuracies are compared theoretically and numerically. The results show
that the new strategy outperforms the current one.Keywords: global sensitivity
analysis, variance-based sensitivity indices, first-order Sobol' index,
total-order Sobol' index, Monte Carlo estimate, asymptotic normality
| stat |
Can Two Forecasts Have the Same Conditional Expected Accuracy? | The method for testing equal predictive accuracy for pairs of forecasting
models proposed by Giacomini and White (2006) has found widespread use in
empirical work. The procedure assumes that the parameters of the underlying
forecasting models are estimated using a rolling window of fixed width and
incorporates the effect of parameter estimation in the null hypothesis that two
forecasts have identical conditionally expected loss. We show that this null
hypothesis cannot be valid under a rolling window estimation scheme and even
fails in the absence of parameter estimation for many types of stochastic
processes in common use. This means that the approach does not guarantee
appropriate comparisons of predictive accuracy of forecasting models. We also
show that the Giacomini-White approach can lead to substantial size distortions
in tests of equal unconditional predictive accuracy and propose an alternative
procedure with better properties.
| stat |
Kernel Estimator and Bandwidth Selection for Density and its
Derivatives: The kedd Package | The kedd package providing additional smoothing techniques to the R
statistical system. Although various packages on the Comprehensive R Archive
Network (CRAN) provide functions useful to nonparametric statistics, kedd aims
to serve as a central location for more specifically of a nonparametric
functions and data sets. The current feature set of the package can be split in
four main categories: compute the convolutions and derivatives of a kernel
function, compute the kernel estimators for a density of probability and its
derivatives, computing the bandwidth selectors with different methods,
displaying the kernel estimators and selection functions of the bandwidth.
Moreover, the package follows the general R philosophy of working with model
objects. This means that instead of merely returning, say, a kernel estimator
of rth derivative of a density, many functions will return an object
containing, it's functions are S3 classes (S3method). The object can then be
manipulated at one's will using various extraction, summary or plotting
functions. Whenever possible, we develop a graphical user interface of the
various functions of a coherent whole, to facilitate the use of this package.
| stat |
Practical Bayesian Optimization of Objectives with Conditioning
Variables | Bayesian optimization is a class of data efficient model based algorithms
typically focused on global optimization. We consider the more general case
where a user is faced with multiple problems that each need to be optimized
conditional on a state variable, for example given a range of cities with
different patient distributions, we optimize the ambulance locations
conditioned on patient distribution. Given partitions of CIFAR-10, we optimize
CNN hyperparameters for each partition. Similarity across objectives boosts
optimization of each objective in two ways: in modelling by data sharing across
objectives, and also in acquisition by quantifying how a single point on one
objective can provide benefit to all objectives. For this we propose a
framework for conditional optimization: ConBO. This can be built on top of a
range of acquisition functions and we propose a new Hybrid Knowledge Gradient
acquisition function. The resulting method is intuitive and theoretically
grounded, performs either similar to or significantly better than recently
published works on a range of problems, and is easily parallelized to collect a
batch of points.
| stat |
Testing biodiversity using inhomogeneous summary statistics | McGill's theory of biodiversity is based upon three axioms: individuals of
the same species cluster together, many rare species co-exist with a few common
ones and individuals of different species grow independently of each other.
Over the past decade, classical point pattern analyses have been employed to
verify these axioms based on the false assumption of stationarity. In this
paper, we use inhomogeneous versions of the classical summary statistics for
spatial point patterns to assess the validity of McGill's first and third
axioms for data obtained from a 50 hectare plot on Barro Colorado Island.
| stat |
Byzantine-Resilient High-Dimensional Federated Learning | We study stochastic gradient descent (SGD) with local iterations in the
presence of malicious/Byzantine clients, motivated by the federated learning.
The clients, instead of communicating with the central server in every
iteration, maintain their local models, which they update by taking several SGD
iterations based on their own datasets and then communicate the net update with
the server, thereby achieving communication-efficiency. Furthermore, only a
subset of clients communicate with the server, and this subset may be different
at different synchronization times. The Byzantine clients may collaborate and
send arbitrary vectors to the server to disrupt the learning process. To combat
the adversary, we employ an efficient high-dimensional robust mean estimation
algorithm from Steinhardt et al.~\cite[ITCS 2018]{Resilience_SCV18} at the
server to filter-out corrupt vectors; and to analyze the outlier-filtering
procedure, we develop a novel matrix concentration result that may be of
independent interest.
We provide convergence analyses for strongly-convex and non-convex smooth
objectives in the heterogeneous data setting, where different clients may have
different local datasets, and we do not make any probabilistic assumptions on
data generation. We believe that ours is the first Byzantine-resilient
algorithm and analysis with local iterations. We derive our convergence results
under minimal assumptions of bounded variance for SGD and bounded gradient
dissimilarity (which captures heterogeneity among local datasets). We also
extend our results to the case when clients compute full-batch gradients.
| stat |
Modeling Interval Trendlines: Symbolic Singular Spectrum Analysis for
Interval Time Series | In this article we propose an extension of singular spectrum analysis for
interval-valued time series. The proposed methods can be used to decompose and
forecast the dynamics governing a set-valued stochastic process. The resulting
components on which the interval time series is decomposed can be understood as
interval trendlines, cycles, or noise. Forecasting can be conducted through a
linear recurrent method, and we devised generalizations of the decomposition
method for the multivariate setting. The performance of the proposed methods is
showcased in a simulation study. We apply the proposed methods so to track the
dynamics governing the Argentina Stock Market (MERVAL) in real time, in a case
study that covers the most recent period of turbulence that led to discussions
of the government of Argentina with the International Monetary Fund.
| stat |
Towards causally interpretable meta-analysis: transporting inferences
from multiple studies to a target population | We take steps towards causally interpretable meta-analysis by describing
methods for transporting causal inferences from a collection of randomized
trials to a new target population, one-trial-at-a-time and pooling all trials.
We discuss identifiability conditions for average treatment effects in the
target population and provide identification results. We show that assuming
inferences are transportable from all trials in the collection to the same
target population has implications for the law underlying the observed data. We
propose average treatment effect estimators that rely on different working
models and provide code for their implementation in statistical software. We
discuss how to use the data to examine whether transported inferences are
homogeneous across the collection of trials, sketch approaches for sensitivity
analysis to violations of the identifiability conditions, and describe
extensions to address non-adherence in the trials. Last, we illustrate the
proposed methods using data from the HALT-C multi-center trial.
| stat |
A Compositional Atlas of Tractable Circuit Operations: From Simple
Transformations to Complex Information-Theoretic Queries | Circuit representations are becoming the lingua franca to express and reason
about tractable generative and discriminative models. In this paper, we show
how complex inference scenarios for these models that commonly arise in machine
learning -- from computing the expectations of decision tree ensembles to
information-theoretic divergences of deep mixture models -- can be represented
in terms of tractable modular operations over circuits. Specifically, we
characterize the tractability of a vocabulary of simple transformations --
sums, products, quotients, powers, logarithms, and exponentials -- in terms of
sufficient structural constraints of the circuits they operate on, and present
novel hardness results for the cases in which these properties are not
satisfied. Building on these operations, we derive a unified framework for
reasoning about tractable models that generalizes several results in the
literature and opens up novel tractable inference scenarios.
| stat |
A positive-definiteness-assured block Gibbs sampler for Bayesian
graphical models with shrinkage priors | Although the block Gibbs sampler for the Bayesian graphical LASSO proposed by
Wang (2012) has been widely applied and extended to various shrinkage priors in
recent years, it has a less noticeable but possibly severe disadvantage that
the positive definiteness of a precision matrix in the Gaussian graphical model
is not guaranteed in each cycle of the Gibbs sampler. Specifically, if the
dimension of the precision matrix exceeds the sample size, the positive
definiteness of the precision matrix will be barely satisfied and the Gibbs
sampler will almost surely fail. In this paper, we propose modifying the
original block Gibbs sampler so that the precision matrix never fails to be
positive definite by sampling it exactly from the domain of the positive
definiteness. As we have shown in the Monte Carlo experiments, this
modification not only stabilizes the sampling procedure but also significantly
improves the performance of the parameter estimation and graphical structure
learning. We also apply our proposed algorithm to a graphical model of the
monthly return data in which the number of stocks exceeds the sample period,
demonstrating its stability and scalability.
| stat |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.