title
stringlengths 6
244
| abstract
stringlengths 19
5.09k
| label
stringclasses 10
values |
|---|---|---|
DRHotNet: An R package for detecting differential risk hotspots on a
linear network | One of the most common applications of spatial data analysis is detecting
zones, at a certain investigation level, where a point-referenced event under
study is especially concentrated. The detection of this kind of zones, which
are usually referred to as hotspots, is essential in certain fields such as
criminology, epidemiology or traffic safety. Traditionally, hotspot detection
procedures have been developed over areal units of analysis. Although working
at this spatial scale can be suitable enough for many research or practical
purposes, detecting hotspots at a more accurate level (for instance, at the
road segment level) may be more convenient sometimes. Furthermore, it is
typical that hotspot detection procedures are entirely focused on the
determination of zones where an event is (overall) highly concentrated. It is
less common, by far, that such procedures prioritize the location of zones
where a specific type of event is overrepresented in relation to the other
types observed, which have been denoted as differential risk hotspots. The R
package DRHotNet provides several functionalities to facilitate the detection
of differential risk hotspots along a linear network. In this paper, DRHotNet
is depicted and its usage in the R console is shown through a detailed analysis
of a crime dataset.
| stat |
The Area of the Convex Hull of Sampled Curves: a Robust Functional
Statistical Depth Measure | With the ubiquity of sensors in the IoT era, statistical observations are
becoming increasingly available in the form of massive (multivariate)
time-series. Formulated as unsupervised anomaly detection tasks, an abundance
of applications like aviation safety management, the health monitoring of
complex infrastructures or fraud detection can now rely on such functional
data, acquired and stored with an ever finer granularity. The concept of
statistical depth, which reflects centrality of an arbitrary observation w.r.t.
a statistical population may play a crucial role in this regard, anomalies
corresponding to observations with 'small' depth. Supported by sound
theoretical and computational developments in the recent decades, it has proven
to be extremely useful, in particular in functional spaces. However, most
approaches documented in the literature consist in evaluating independently the
centrality of each point forming the time series and consequently exhibit a
certain insensitivity to possible shape changes. In this paper, we propose a
novel notion of functional depth based on the area of the convex hull of
sampled curves, capturing gradual departures from centrality, even beyond the
envelope of the data, in a natural fashion. We discuss practical relevance of
commonly imposed axioms on functional depths and investigate which of them are
satisfied by the notion of depth we promote here. Estimation and computational
issues are also addressed and various numerical experiments provide empirical
evidence of the relevance of the approach proposed.
| stat |
Exponentially-Modified Gaussian Mixture Model: Applications in
Spectroscopy | We propose a novel exponentially-modified Gaussian (EMG) mixture residual
model. The EMG mixture is well suited to model residuals that are contaminated
by a distribution with positive support. This is in contrast to commonly used
robust residual models, like the Huber loss or $\ell_1$, which assume a
symmetric contaminating distribution and are otherwise asymptotically biased.
We propose an expectation-maximization algorithm to optimize an arbitrary model
with respect to the EMG mixture. We apply the approach to linear regression and
probabilistic matrix factorization (PMF). We compare against other residual
models, including quantile regression. Our numerical experiments demonstrate
the strengths of the EMG mixture on both tasks. The PMF model arises from
considering spectroscopic data. In particular, we demonstrate the effectiveness
of PMF in conjunction with the EMG mixture model on synthetic data and two
real-world applications: X-ray diffraction and Raman spectroscopy. We show how
our approach is effective in inferring background signals and systematic errors
in data arising from these experimental settings, dramatically outperforming
existing approaches and revealing the data's physically meaningful components.
| stat |
Statistical approaches using longitudinal biomarkers for disease early
detection: A comparison of methodologies | Early detection of clinical outcomes such as cancer may be predicted based on
longitudinal biomarker measurements. Tracking longitudinal biomarkers as a way
to identify early disease onset may help to reduce mortality from diseases like
ovarian cancer that are more treatable if detected early. Two general
frameworks for disease risk prediction, the shared random effects model (SREM)
and the pattern mixture model (PMM) could be used to assess longitudinal
biomarkers on disease early detection. In this paper, we studied the predictive
performances of SREM and PMM on disease early detection through an application
to ovarian cancer, where early detection using the risk of ovarian cancer
algorithm (ROCA) has been evaluated. Comparisons of the above three methods
were performed via the analyses of the ovarian cancer data from the Prostate,
Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial and extensive
simulation studies. The time-dependent receiving operating characteristic (ROC)
curve and its area (AUC) were used to evaluate the prediction accuracy. The
out-of-sample predictive performance was calculated using leave-one-out
cross-validation (LOOCV), aiming to minimize the problem of model over-fitting.
A careful analysis of the use of the biomarker cancer antigen 125 for ovarian
cancer early detection showed improved performance of PMM as compared with SREM
and ROCA. More generally, simulation studies showed that PMM outperforms ROCA
unless biomarkers are taken at very frequent screening settings.
| stat |
History matching with probabilistic emulators and active learning | The scientific understanding of real-world processes has dramatically
improved over the years through computer simulations. Such simulators represent
complex mathematical models that are implemented as computer codes which are
often expensive. The validity of using a particular simulator to draw accurate
conclusions relies on the assumption that the computer code is correctly
calibrated. This calibration procedure is often pursued under extensive
experimentation and comparison with data from a real-world process. The problem
is that the data collection may be so expensive that only a handful of
experiments are feasible. History matching is a calibration technique that,
given a simulator, it iteratively discards regions of the input space using an
implausibility measure. When the simulator is computationally expensive, an
emulator is used to explore the input space. In this paper, a Gaussian process
provides a complete probabilistic output that is incorporated into the
implausibility measure. The identification of regions of interest is
accomplished with recently developed annealing sampling techniques. Active
learning functions are incorporated into the history matching procedure to
refocus on the input space and improve the emulator. The efficiency of the
proposed framework is tested in well-known examples from the history matching
literature, as well as in a proposed testbed of functions of higher dimensions.
| stat |
Kalman Filter, Sensor Fusion, and Constrained Regression: Equivalences
and Insights | The Kalman filter (KF) is one of the most widely used tools for data
assimilation and sequential estimation. In this work, we show that the state
estimates from the KF in a standard linear dynamical system setting are
equivalent to those given by the KF in a transformed system, with infinite
process noise (i.e., a "flat prior") and an augmented measurement space. This
reformulation---which we refer to as augmented measurement sensor fusion
(SF)---is conceptually interesting, because the transformed system here is
seemingly static (as there is effectively no process model), but we can still
capture the state dynamics inherent to the KF by folding the process model into
the measurement space. Further, this reformulation of the KF turns out to be
useful in settings in which past states are observed eventually (at some lag).
Here, when the measurement noise covariance is estimated by the empirical
covariance, we show that the state predictions from SF are equivalent to those
from a regression of past states on past measurements, subject to particular
linear constraints (reflecting the relationships encoded in the measurement
map). This allows us to port standard ideas (say, regularization methods) in
regression over to dynamical systems. For example, we can posit multiple
candidate process models, fold all of them into the measurement model,
transform to the regression perspective, and apply $\ell_1$ penalization to
perform process model selection. We give various empirical demonstrations, and
focus on an application to nowcasting the weekly incidence of influenza in the
US.
| stat |
Independence Tests Without Ground Truth for Noisy Learners | Exact ground truth invariant polynomial systems can be written for
arbitrarily correlated binary classifiers. Their solutions give estimates for
sample statistics that require knowledge of the ground truth of the correct
labels in the sample. Of these polynomial systems, only a few have been solved
in closed form. Here we discuss the exact solution for independent binary
classifiers - resolving an outstanding problem that has been presented at this
conference and others. Its practical applicability is hampered by its sole
remaining assumption - the classifiers need to be independent in their sample
errors. We discuss how to use the closed form solution to create a
self-consistent test that can validate the independence assumption itself
absent the correct labels ground truth. It can be cast as an algebraic geometry
conjecture for binary classifiers that remains unsolved. A similar conjecture
for the ground truth invariant algebraic system for scalar regressors is
solvable, and we present the solution here. We also discuss experiments on the
Penn ML Benchmark classification tasks that provide further evidence that the
conjecture may be true for the polynomial system of binary classifiers.
| stat |
Curating a COVID-19 data repository and forecasting county-level death
counts in the United States | As the COVID-19 outbreak evolves, accurate forecasting continues to play an
extremely important role in informing policy decisions. In this paper, we
present our continuous curation of a large data repository containing COVID-19
information from a range of sources. We use this data to develop predictions
and corresponding prediction intervals for the short-term trajectory of
COVID-19 cumulative death counts at the county-level in the United States up to
two weeks ahead. Using data from January 22 to June 20, 2020, we develop and
combine multiple forecasts using ensembling techniques, resulting in an
ensemble we refer to as Combined Linear and Exponential Predictors (CLEP). Our
individual predictors include county-specific exponential and linear
predictors, a shared exponential predictor that pools data together across
counties, an expanded shared exponential predictor that uses data from
neighboring counties, and a demographics-based shared exponential predictor. We
use prediction errors from the past five days to assess the uncertainty of our
death predictions, resulting in generally-applicable prediction intervals,
Maximum (absolute) Error Prediction Intervals (MEPI). MEPI achieves a coverage
rate of more than 94% when averaged across counties for predicting cumulative
recorded death counts two weeks in the future. Our forecasts are currently
being used by the non-profit organization, Response4Life, to determine the
medical supply need for individual hospitals and have directly contributed to
the distribution of medical supplies across the country. We hope that our
forecasts and data repository at https://covidseverity.com can help guide
necessary county-specific decision-making and help counties prepare for their
continued fight against COVID-19.
| stat |
From predictions to prescriptions: A data-driven response to COVID-19 | The COVID-19 pandemic has created unprecedented challenges worldwide.
Strained healthcare providers make difficult decisions on patient triage,
treatment and care management on a daily basis. Policy makers have imposed
social distancing measures to slow the disease, at a steep economic price. We
design analytical tools to support these decisions and combat the pandemic.
Specifically, we propose a comprehensive data-driven approach to understand the
clinical characteristics of COVID-19, predict its mortality, forecast its
evolution, and ultimately alleviate its impact. By leveraging cohort-level
clinical data, patient-level hospital data, and census-level epidemiological
data, we develop an integrated four-step approach, combining descriptive,
predictive and prescriptive analytics. First, we aggregate hundreds of clinical
studies into the most comprehensive database on COVID-19 to paint a new
macroscopic picture of the disease. Second, we build personalized calculators
to predict the risk of infection and mortality as a function of demographics,
symptoms, comorbidities, and lab values. Third, we develop a novel
epidemiological model to project the pandemic's spread and inform social
distancing policies. Fourth, we propose an optimization model to re-allocate
ventilators and alleviate shortages. Our results have been used at the clinical
level by several hospitals to triage patients, guide care management, plan ICU
capacity, and re-distribute ventilators. At the policy level, they are
currently supporting safe back-to-work policies at a major institution and
equitable vaccine distribution planning at a major pharmaceutical company, and
have been integrated into the US Center for Disease Control's pandemic
forecast.
| stat |
A multiple-frame approach of crop yield estimation from satellite
remotely sensed data | Many studies have recently explored the information from the
satellite-remotely sensed data (SRSD) for estimating the crop production
statistics. The value of this information depends on the aerial and spatial
resolutions of SRSD. The SRSD with fine spatial resolution is costly and the
aerial coverage is less. Use of multiple frames of SRSD in the estimation
process of crop production can increase the precision. We propose an estimator
for the average yield of wheat for the state of Haryana, India. This estimator
uses the information from the Wide Field Sensor (WiFS) and the Linear Imaging
Self Scanner (LISS-III) data from the Indian Remote Sensing satellite (IRS-1D)
and the crop cutting experiment data collected by probability sampling design
from a list frame of villages. We find that the relative efficiencies of the
multiple-frame estimators are high in comparison to the single frame
estimators.
| stat |
Physics Regularized Gaussian Processes | We consider incorporating incomplete physics knowledge, expressed as
differential equations with latent functions, into Gaussian processes (GPs) to
improve their performance, especially for limited data and extrapolation. While
existing works have successfully encoded such knowledge via kernel convolution,
they only apply to linear equations with analytical Green's functions. The
convolution can further restrict us from fusing physics with highly expressive
kernels, e.g., deep kernels. To overcome these limitations, we propose Physics
Regularized Gaussian Process (PRGP) that can incorporate both linear and
nonlinear equations, does not rely on Green's functions, and is free to use
arbitrary kernels. Specifically, we integrate the standard GP with a generative
model to encode the differential equation in a principled Bayesian hybrid
framework. For efficient and effective inference, we marginalize out the latent
variables and derive a simplified model evidence lower bound (ELBO), based on
which we develop a stochastic collapsed inference algorithm. Our ELBO can be
viewed as a posterior regularization objective. We show the advantage of our
approach in both simulation and real-world applications.
| stat |
Accounting for not-at-random missingness through imputation stacking | Not-at-random missingness presents a challenge in addressing missing data in
many health research applications. In this paper, we propose a new approach to
account for not-at-random missingness after multiple imputation through
weighted analysis of stacked multiple imputations. The weights are easily
calculated as a function of the imputed data and assumptions about the
not-at-random missingness. We demonstrate through simulation that the proposed
method has excellent performance when the missingness model is correctly
specified. In practice, the missingness mechanism will not be known. We show
how we can use our approach in a sensitivity analysis framework to evaluate the
robustness of model inference to different assumptions about the missingness
mechanism, and we provide R package StackImpute to facilitate implementation as
part of routine sensitivity analyses. We apply the proposed method to account
for not-at-random missingness in human papillomavirus test results in a study
of survival for patients diagnosed with oropharyngeal cancer.
| stat |
Hierarchical network design for nitrogen dioxide measurement in urban
environments, part 1: proxy selection | Previous studies have shown that a hierarchical network comprising a number
of compliant reference stations and a much larger number of low-cost sensors
can deliver reliable air quality data at high temporal and spatial resolution
for ozone at neighbourhood scales. Key to this framework is the concept of a
proxy: a reliable (regulatory) data source whose results have sufficient
statistical similarity over some period of time to those from any given
low-cost measurement site. This enables the low-cost instruments to be
calibrated remotely, avoiding the need for costly on-site calibration of dense
networks. This paper assesses the suitability of this method for local air
pollutants such as nitrogen dioxide which show large temporal and spatial
variability in concentration. The proxy technique is evaluated using the data
from the network of regulatory air monitoring stations measuring nitrogen
dioxide in Southern California to avoid errors introduced by low-cost
instrument performance. Proxies chosen based on land use similarity signalled
typically less than 0.1 percent false alarms. Although poor proxy performance
was observed when the local geography was unusual (a semi-enclosed valley) in
this instance the closest neighbour station proved to be an appropriate
alternative. The method also struggled when wind speeds were low and very local
sources presumably dominated the concentration patterns. Overall, we
demonstrate that the technique can be applied to nitrogen dioxide, and that
appropriate proxies can be found even within a spatially sparse network of
stations in a region with large spatio-temporal variation in concentration.
| stat |
The bivariate $K$-finite normal mixture "blanket" copula: an application
to driving patterns | There are many bivariate parametric copulas in the literature to model
bivariate data with different dependence features. We propose a new bivariate
parametric copula family that cannot only handle various dependence patterns
that appear in the existing parametric bivariate copula families, but also
provides a more enriched dependence structure. The proposed copula construction
exploits finite mixtures of bivariate normal distributions. The mixing
operation, the distinct correlation and mean parameters at each mixture
component introduce quite a flexible dependence. We apply the new copula to
real transportation data that cannot apparently be modelled by any of the
existing parametric families of bivariate copulas.
| stat |
Practical Hilbert space approximate Bayesian Gaussian processes for
probabilistic programming | Gaussian processes are powerful non-parametric probabilistic models for
stochastic functions. However they entail a complexity that is computationally
intractable when the number of observations is large, especially when estimated
with fully Bayesian methods such as Markov chain Monte Carlo. In this paper, we
focus on a novel approach for low-rank approximate Bayesian Gaussian processes,
based on a basis function approximation via Laplace eigenfunctions for
stationary covariance functions. The main contribution of this paper is a
detailed analysis of the performance and practical implementation of the method
in relation to key factors such as the number of basis functions, domain of the
prediction space, and smoothness of the latent function. We provide intuitive
visualizations and recommendations for choosing the values of these factors,
which make it easier for users to improve approximation accuracy and
computational performance. We also propose diagnostics for checking that the
number of basis functions and the domain of the prediction space are adequate
given the data. The proposed approach is simple and exhibits an attractive
computational complexity due to its linear structure, and it is easy to
implement in probabilistic programming frameworks. Several illustrative
examples of the performance and applicability of the method in the
probabilistic programming language Stan are presented together with the
underlying Stan model code.
| stat |
Evaluation of Ride-Sourcing Search Frictions and Driver Productivity: A
Spatial Denoising Approach | This paper considers the problem of measuring spatial and temporal variation
in driver productivity on ride-sourcing trips. This variation is especially
important from a driver's perspective: if a platform's drivers experience
systematic disparities in earnings because of variation in their riders'
destinations, they may perceive the pricing model as inequitable. This
perception can exacerbate search frictions if it leads drivers to avoid
locations where they believe they may be assigned "unlucky" fares. To
characterize any such systematic disparities in productivity, we develop an
analytic framework with three key components. First, we propose a productivity
metric that looks two consecutive trips ahead, thus capturing the effect on
expected earnings of market conditions at drivers' drop-off locations. Second,
we develop a natural experiment by analyzing trips with a common origin but
varying destinations, thus isolating purely spatial effects on productivity.
Third, we apply a spatial denoising method that allows us to work with raw
spatial information exhibiting high levels of noise and sparsity, without
having to aggregate data into large, low-resolution spatial zones. By applying
our framework to data on more than 1.4 million rides in Austin, Texas, we find
significant spatial variation in ride-sourcing driver productivity and search
frictions. Drivers at the same location experienced disparities in productivity
after being dispatched on trips with different destinations, with origin-based
surge pricing increasing these earnings disparities. Our results show that trip
distance is the dominant factor in driver productivity: short trips yielded
lower productivity, even when ending in areas with high demand. These findings
suggest that new pricing strategies are required to minimize random disparities
in driver earnings.
| stat |
Learning Deep Analysis Dictionaries -- Part II: Convolutional
Dictionaries | In this paper, we introduce a Deep Convolutional Analysis Dictionary Model
(DeepCAM) by learning convolutional dictionaries instead of unstructured
dictionaries as in the case of deep analysis dictionary model introduced in the
companion paper. Convolutional dictionaries are more suitable for processing
high-dimensional signals like for example images and have only a small number
of free parameters. By exploiting the properties of a convolutional dictionary,
we present an efficient convolutional analysis dictionary learning approach. A
L-layer DeepCAM consists of L layers of convolutional analysis dictionary and
element-wise soft-thresholding pairs and a single layer of convolutional
synthesis dictionary. Similar to DeepAM, each convolutional analysis dictionary
is composed of a convolutional Information Preserving Analysis Dictionary
(IPAD) and a convolutional Clustering Analysis Dictionary (CAD). The IPAD and
the CAD are learned using variations of the proposed learning algorithm. We
demonstrate that DeepCAM is an effective multilayer convolutional model and, on
single image super-resolution, achieves performance comparable with other
methods while also showing good generalization capabilities.
| stat |
Functional Time Series Forecasting: Functional Singular Spectrum
Analysis Approaches | In this paper, we propose two nonparametric methods used in the forecasting
of functional time-dependent data, namely functional singular spectrum analysis
recurrent forecasting and vector forecasting. Both algorithms utilize the
results of functional singular spectrum analysis and past observations in order
to predict future data points where recurrent forecasting predicts one function
at a time and the vector forecasting makes predictions using functional
vectors. We compare our forecasting methods to a gold standard algorithm used
in the prediction of functional, time-dependent data by way of simulation and
real data and we find our techniques do better for periodic stochastic
processes.
| stat |
A Causal Bayesian Networks Viewpoint on Fairness | We offer a graphical interpretation of unfairness in a dataset as the
presence of an unfair causal path in the causal Bayesian network representing
the data-generation mechanism. We use this viewpoint to revisit the recent
debate surrounding the COMPAS pretrial risk assessment tool and, more
generally, to point out that fairness evaluation on a model requires careful
considerations on the patterns of unfairness underlying the training data. We
show that causal Bayesian networks provide us with a powerful tool to measure
unfairness in a dataset and to design fair models in complex unfairness
scenarios.
| stat |
Data-based wind disaster climate identification algorithm and extreme
wind speed prediction | An extreme wind speed estimation method that considers wind hazard climate
types is critical for design wind load calculation for building structures
affected by mixed climates. However, it is very difficult to obtain wind hazard
climate types from meteorological data records, because they restrict the
application of extreme wind speed estimation in mixed climates. This paper
first proposes a wind hazard type identification algorithm based on a numerical
pattern recognition method that utilizes feature extraction and generalization.
Next, it compares six commonly used machine learning models using K-fold
cross-validation. Finally, it takes meteorological data from three locations
near the southeast coast of China as examples to examine the algorithm
performance. Based on classification results, the extreme wind speeds
calculated based on mixed wind hazard types is compared with those obtained
from conventional methods, and the effects on structural design for different
return periods are discussed.
| stat |
Learning to Rank with Missing Data via Generative Adversarial Networks | We explore the role of Conditional Generative Adversarial Networks (GAN) in
imputing missing data and apply GAN imputation on a novel use case in
e-commerce: a learning-to-rank problem with incomplete training data.
Conventional imputation methods often make assumptions regarding the underlying
distribution of the missing data, while GANs offer an alternative framework to
sidestep approximating intractable distributions. First, we prove that GAN
imputation offers theoretical guarantees beyond the naive Missing Completely At
Random (MCAR) scenario. Next, we show that empirically, the Conditional GAN
structure is well suited for data with heterogeneous distributions and across
unbalanced classes, improving performance metrics such as RMSE. Using an Amazon
Search ranking dataset, we produce standard ranking models trained on
GAN-imputed data that are comparable to training on ground-truth data based on
standard ranking quality metrics NDCG and MRR. We also highlight how different
neural net features such as convolution and dropout layers can improve
performance given different missing value settings.
| stat |
Optimal tournament design: lessons from the men's handball Champions
League | Many sports tournaments are organised in a hybrid design consisting of a
round-robin group stage followed by a knock-out phase. The traditional seeding
regime aims to create balanced groups roughly at the same competition level but
may result in several uneven matches when the quality of the teams varies
greatly. Our paper is the first challenging this classical solution through the
example of the men's EHF (European Handball Federation) Champions League, the
most prestigious men's handball club competition in Europe, which has used
unbalanced groups between the 2015/16 and 2019/20 seasons. Its particular
design is compared to an alternative format with equally strong groups, as well
as to the previous scheme of the EHF Champions League. We find that it is
possible to increase the quality of all matches played together with raising
the uncertainty of outcome, essentially without sacrificing fairness. Our
results have useful implications for the governing bodies of major sports.
| stat |
Parallelizing MCMC Sampling via Space Partitioning | Efficient sampling of many-dimensional and multimodal density functions is a
task of great interest in many research fields. We describe an algorithm that
allows parallelizing inherently serial Markov chain Monte Carlo (MCMC) sampling
by partitioning the space of the function parameters into multiple subspaces
and sampling each of them independently. The samples of the different subspaces
are then reweighted by their integral values and stitched back together. This
approach allows reducing sampling wall-clock time by parallel operation. It
also improves sampling of multimodal target densities and results in less
correlated samples. Finally, the approach yields an estimate of the integral of
the target density function.
| stat |
Deep Learning for Survival Outcomes | This manuscripts develops a new class of deep learning algorithms for
outcomes that are potentially censored. To account for censoring, the
unobservable loss function used in the absence of censoring is replaced by a
censoring unbiased transformation. The resulting class of algorithms can be
used to estimate both survival probabilities and restricted mean survival. We
show how the deep learning algorithms can be implemented using software for
uncensored data using a form of response transformation. Simulations and
analysis of the Netherlands 70 Gene Signature Data show strong performance of
the proposed algorithms.
| stat |
Walsh-Hadamard Variational Inference for Bayesian Deep Learning | Over-parameterized models, such as DeepNets and ConvNets, form a class of
models that are routinely adopted in a wide variety of applications, and for
which Bayesian inference is desirable but extremely challenging. Variational
inference offers the tools to tackle this challenge in a scalable way and with
some degree of flexibility on the approximation, but for over-parameterized
models this is challenging due to the over-regularization property of the
variational objective. Inspired by the literature on kernel methods, and in
particular on structured approximations of distributions of random matrices,
this paper proposes Walsh-Hadamard Variational Inference (WHVI), which uses
Walsh-Hadamard-based factorization strategies to reduce the parameterization
and accelerate computations, thus avoiding over-regularization issues with the
variational objective. Extensive theoretical and empirical analyses demonstrate
that WHVI yields considerable speedups and model reductions compared to other
techniques to carry out approximate inference for over-parameterized models,
and ultimately show how advances in kernel methods can be translated into
advances in approximate Bayesian inference.
| stat |
A unified method for multivariate mixed-type response regression | We propose a new method for multivariate response regressions where the
elements of the response vector can be of mixed types, for example some
continuous and some discrete. Our method is based on a model which assumes the
observable mixed-type response vector is connected to a latent multivariate
normal response linear regression through a link function. We explore the
properties of this model and show its parameters are identifiable under
reasonable conditions. We propose an algorithm for approximate maximum
likelihood estimation that works "off-the-shelf" with many different
combinations of response types, and which scales well in the dimension of the
response vector. Our method typically gives better predictions and parameter
estimates than fitting separate models for the different response types and
allows for approximate likelihood ratio testing of relevant hypotheses such as
independence of responses. The usefulness of the proposed method is illustrated
using simulations and through three data examples.
| stat |
Unreliable Multi-Armed Bandits: A Novel Approach to Recommendation
Systems | We use a novel modification of Multi-Armed Bandits to create a new model for
recommendation systems. We model the recommendation system as a bandit seeking
to maximize reward by pulling on arms with unknown rewards. The catch however
is that this bandit can only access these arms through an unreliable
intermediate that has some level of autonomy while choosing its arms. For
example, in a streaming website the user has a lot of autonomy while choosing
content they want to watch. The streaming sites can use targeted advertising as
a means to bias opinions of these users. Here the streaming site is the bandit
aiming to maximize reward and the user is the unreliable intermediate. We model
the intermediate as accessing states via a Markov chain. The bandit is allowed
to perturb this Markov chain. We prove fundamental theorems for this setting
after which we show a close-to-optimal Explore-Commit algorithm.
| stat |
Direct Feedback Alignment Scales to Modern Deep Learning Tasks and
Architectures | Despite being the workhorse of deep learning, the backpropagation algorithm
is no panacea. It enforces sequential layer updates, thus preventing efficient
parallelization of the training process. Furthermore, its biological
plausibility is being challenged. Alternative schemes have been devised; yet,
under the constraint of synaptic asymmetry, none have scaled to modern deep
learning tasks and architectures. Here, we challenge this perspective, and
study the applicability of Direct Feedback Alignment to neural view synthesis,
recommender systems, geometric learning, and natural language processing. In
contrast with previous studies limited to computer vision tasks, our findings
show that it successfully trains a large range of state-of-the-art deep
learning architectures, with performance close to fine-tuned backpropagation.
At variance with common beliefs, our work supports that challenging tasks can
be tackled in the absence of weight transport.
| stat |
Pearson chi^2-divergence Approach to Gaussian Mixture Reduction and its
Application to Gaussian-sum Filter and Smoother | The Gaussian mixture distribution is important in various statistical
problems. In particular it is used in the Gaussian-sum filter and smoother for
linear state-space model with non-Gaussian noise inputs. However, for this
method to be practical, an efficient method of reducing the number of Gaussian
components is necessary. In this paper, we show that a closed form expression
of Pearson chi^2-divergence can be obtained and it can apply to the
determination of the pair of two Gaussian components in sequential reduction of
Gaussian components. By numerical examples for one dimensional and two
dimensional distribution models, it will be shown that in most cases the
proposed criterion performed almost equally as the Kullback-Libler divergence,
for which computationally costly numerical integration is necessary.
Application to Gaussian-sum filtering and smoothing is also shown.
| stat |
Chromatic Learning for Sparse Datasets | Learning over sparse, high-dimensional data frequently necessitates the use
of specialized methods such as the hashing trick. In this work, we design a
highly scalable alternative approach that leverages the low degree of feature
co-occurrences present in many practical settings. This approach, which we call
Chromatic Learning (CL), obtains a low-dimensional dense feature representation
by performing graph coloring over the co-occurrence graph of features---an
approach previously used as a runtime performance optimization for GBDT
training. This color-based dense representation can be combined with additional
dense categorical encoding approaches, e.g., submodular feature compression, to
further reduce dimensionality. CL exhibits linear parallelizability and
consumes memory linear in the size of the co-occurrence graph. By leveraging
the structural properties of the co-occurrence graph, CL can compress sparse
datasets, such as KDD Cup 2012, that contain over 50M features down to 1024,
using an order of magnitude fewer features than frequency-based truncation and
the hashing trick while maintaining the same test error for linear models. This
compression further enables the use of deep networks in this wide, sparse
setting, where CL similarly has favorable performance compared to existing
baselines for budgeted input dimension.
| stat |
A Critique of Differential Abundance Analysis, and Advocacy for an
Alternative | It is largely taken for granted that differential abundance analysis is, by
default, the best first step when analyzing genomic data. We argue that this is
not necessarily the case. In this article, we identify key limitations that are
intrinsic to differential abundance analysis: it is (a) dependent on
unverifiable assumptions, (b) an unreliable construct, and (c) overly
reductionist. We formulate an alternative framework called ratio-based
biomarker analysis which does not suffer from the identified limitations.
Moreover, ratio-based biomarkers are highly flexible. Beyond replacing DAA,
they can also be used for many other bespoke analyses, including dimension
reduction and multi-omics data integration.
| stat |
Concept Drift Detection and Adaptation with Hierarchical Hypothesis
Testing | A fundamental issue for statistical classification models in a streaming
environment is that the joint distribution between predictor and response
variables changes over time (a phenomenon also known as concept drifts), such
that their classification performance deteriorates dramatically. In this paper,
we first present a hierarchical hypothesis testing (HHT) framework that can
detect and also adapt to various concept drift types (e.g., recurrent or
irregular, gradual or abrupt), even in the presence of imbalanced data labels.
A novel concept drift detector, namely Hierarchical Linear Four Rates (HLFR),
is implemented under the HHT framework thereafter. By substituting a
widely-acknowledged retraining scheme with an adaptive training strategy, we
further demonstrate that the concept drift adaptation capability of HLFR can be
significantly boosted. The theoretical analysis on the Type-I and Type-II
errors of HLFR is also performed. Experiments on both simulated and real-world
datasets illustrate that our methods outperform state-of-the-art methods in
terms of detection precision, detection delay as well as the adaptability
across different concept drift types.
| stat |
Simple statistical models and sequential deep learning for Lithium-ion
batteries degradation under dynamic conditions: Fractional Polynomials vs
Neural Networks | Longevity and safety of Lithium-ion batteries are facilitated by efficient
monitoring and adjustment of the battery operating conditions: hence, it is
crucial to implement fast and accurate algorithms for State of Health (SoH)
monitoring on the Battery Management System. The task is challenging due to the
complexity and multitude of the factors contributing to the battery capacity
degradation, especially because the different degradation processes occur at
various timescales and their interactions play an important role. This paper
proposes and compares two data-driven approaches: a Long Short-Term Memory
neural network, from the field of deep learning, and a Multivariable Fractional
Polynomial regression, from classical statistics. Models from both classes are
trained from historical data of one exhausted cell and used to predict the SoH
of other cells. This work uses data provided by the NASA Ames Prognostics
Center of Excellence, characterised by varying loads which simulate dynamic
operating conditions. Two hypothetical scenarios are considered: one assumes
that a recent true capacity measurement is known, the other relies solely on
the cell nominal capacity. Both methods are effective, with low prediction
errors, and the advantages of one over the other in terms of interpretability
and complexity are discussed in a critical way.
| stat |
Investigation of Patient-sharing Networks Using a Bayesian Network Model
Selection Approach for Congruence Class Models | A Bayesian approach to conduct network model selection is presented for a
general class of network models referred to as the congruence class models
(CCMs). CCMs form a broad class that includes as special cases several common
network models, such as the Erd\H{o}s-R\'{e}nyi-Gilbert model, stochastic block
model and many exponential random graph models. Due to the range of models able
to be specified as a CCM, investigators are better able to select a model
consistent with generative mechanisms associated with the observed network
compared to current approaches. In addition, the approach allows for
incorporation of prior information. We utilize the proposed Bayesian network
model selection approach for CCMs to investigate several mechanisms that may be
responsible for the structure of patient-sharing networks, which are associated
with the cost and quality of medical care. We found evidence in support of
heterogeneity in sociality but not selective mixing by provider type nor
degree.
| stat |
Super-resolution estimation of cyclic arrival rates | Exploiting the fact that most arrival processes exhibit cyclic behaviour, we
propose a simple procedure for estimating the intensity of a nonhomogeneous
Poisson process. The estimator is the super-resolution analogue to Shao 2010
and Shao & Lii 2011, which is a sum of $p$ sinusoids where $p$ and the
frequency, amplitude, and phase of each wave are not known and need to be
estimated. This results in an interpretable yet flexible specification that is
suitable for use in modelling as well as in high resolution simulations.
Our estimation procedure sits in between classic periodogram methods and
atomic/total variation norm thresholding. Through a novel use of window
functions in the point process domain, our approach attains super-resolution
without semidefinite programming. Under suitable conditions, finite sample
guarantees can be derived for our procedure. These resolve some open questions
and expand existing results in spectral estimation literature.
| stat |
Capped norm linear discriminant analysis and its applications | Classical linear discriminant analysis (LDA) is based on squared Frobenious
norm and hence is sensitive to outliers and noise. To improve the robustness of
LDA, in this paper, we introduce capped l_{2,1}-norm of a matrix, which employs
non-squared l_2-norm and "capped" operation, and further propose a novel capped
l_{2,1}-norm linear discriminant analysis, called CLDA. Due to the use of
capped l_{2,1}-norm, CLDA can effectively remove extreme outliers and suppress
the effect of noise data. In fact, CLDA can be also viewed as a weighted LDA.
CLDA is solved through a series of generalized eigenvalue problems with
theoretical convergency. The experimental results on an artificial data set,
some UCI data sets and two image data sets demonstrate the effectiveness of
CLDA.
| stat |
Quasi-maximum Likelihood Inference for Linear Double Autoregressive
Models | This paper investigates the quasi-maximum likelihood inference including
estimation, model selection and diagnostic checking for linear double
autoregressive (DAR) models, where all asymptotic properties are established
under only fractional moment of the observed process. We propose a Gaussian
quasi-maximum likelihood estimator (G-QMLE) and an exponential quasi-maximum
likelihood estimator (E-QMLE) for the linear DAR model, and establish the
consistency and asymptotic normality for both estimators. Based on the G-QMLE
and E-QMLE, two Bayesian information criteria are proposed for model selection,
and two mixed portmanteau tests are constructed to check the adequacy of fitted
models. Moreover, we compare the proposed G-QMLE and E-QMLE with the existing
doubly weighted quantile regression estimator in terms of the asymptotic
efficiency and numerical performance. Simulation studies illustrate the
finite-sample performance of the proposed inference tools, and a real example
on the Bitcoin return series shows the usefulness of the proposed inference
tools.
| stat |
Visualization tools for parameter selection in cluster analysis | We propose an algorithm, HPREF (Hierarchical Partitioning by Repeated
Features), that produces a hierarchical partition of a set of clusterings of a
fixed dataset, such as sets of clusterings produced by running a clustering
algorithm with a range of parameters. This gives geometric structure to such
sets of clustering, and can be used to visualize the set of results one obtains
by running a clustering algorithm with a range of parameters.
| stat |
EigenGame Unloaded: When playing games is better than optimizing | We build on the recently proposed EigenGame that views eigendecomposition as
a competitive game. EigenGame's updates are biased if computed using
minibatches of data, which hinders convergence and more sophisticated
parallelism in the stochastic setting. In this work, we propose an unbiased
stochastic update that is asymptotically equivalent to EigenGame, enjoys
greater parallelism allowing computation on datasets of larger sample sizes,
and outperforms EigenGame in experiments. We present applications to finding
the principal components of massive datasets and performing spectral clustering
of graphs. We analyze and discuss our proposed update in the context of
EigenGame and the shift in perspective from optimization to games.
| stat |
Sparse Network Modeling | There have been many attempts to identify high-dimensional network features
via multivariate approaches. Specifically, when the number of voxels or nodes,
denoted as p, are substantially larger than the number of images, denoted as n,
it produces an under-determined model with infinitely many possible solutions.
The small-n large-p problem is often remedied by regularizing the
under-determined system with additional sparse penalties. Popular sparse
network models include sparse correlations, LASSO, sparse canonical
correlations and graphical-LASSO. These popular sparse models require
optimizing L1-norm penalties, which has been the major computational bottleneck
for solving large-scale problems. Thus, many existing sparse brain network
models in brain imaging have been restricted to a few hundreds nodes or less.
2527 MRI features used in a LASSO model for Alzheimer's disease is probably the
largest number of features used in any sparse model in the brain imaging
literature.
| stat |
Bayesian bandwidth estimation for local linear fitting in nonparametric
regression models | This paper presents a Bayesian sampling approach to bandwidth estimation for
the local linear estimator of the regression function in a nonparametric
regression model. In the Bayesian sampling approach, the error density is
approximated by a location-mixture density of Gaussian densities with means the
individual errors and variance a constant parameter. This mixture density has
the form of a kernel density estimator of errors and is referred to as the
kernel-form error density (c.f., Zhang et al., 2014). While Zhang et al. (2014)
use the local constant (also known as the Nadaraya- Watson) estimator to
estimate the regression function, we extend this to the local linear estimator,
which produces more accurate estimation. The proposed investigation is
motivated by the lack of data-driven methods for simultaneously choosing
bandwidths in the local linear estimator of the regression function and
kernel-form error density. Treating bandwidths as parameters, we derive an
approximate (pseudo) likelihood and a posterior. A simulation study shows that
the proposed bandwidth estimation outperforms the rule-of-thumb and
cross-validation methods under the criterion of integrated squared errors. The
proposed bandwidth estimation method is validated through a nonparametric
regression model involving firm ownership concentration, and a model involving
state-price density estimation.
| stat |
Multiple imputation and selection of ordinal level 2 predictors in
multilevel models. An analysis of the relationship between student ratings
and teacher beliefs and practices | The paper is motivated by the analysis of the relationship between ratings
and teacher practices and beliefs, which are measured via a set of binary and
ordinal items collected by a specific survey with nearly half missing
respondents. The analysis, which is based on a two-level random effect model,
must face two about the items measuring teacher practices and beliefs: (i)
these items level 2 predictors severely affected by missingness; (ii) there is
redundancy in the number of items and the number of categories of their
measurement scale. tackle the first issue by considering a multiple imputation
strategy based on information at both level 1 and level 2. For the second
issue, we consider regularization techniques for ordinal predictors, also
accounting for the multilevel data structure. The proposed solution combines
existing methods in an original way to solve specific problem at hand, but it
is generally applicable to settings requiring to select predictors affected by
missing values. The results obtained with the final model out that some teacher
practices and beliefs are significantly related to ratings about teacher
ability to motivate students.
| stat |
Multivariate Gaussian and Student$-t$ Process Regression for
Multi-output Prediction | Gaussian process model for vector-valued function has been shown to be useful
for multi-output prediction. The existing method for this model is to
re-formulate the matrix-variate Gaussian distribution as a multivariate normal
distribution. Although it is effective in many cases, re-formulation is not
always workable and is difficult to apply to other distributions because not
all matrix-variate distributions can be transformed to respective multivariate
distributions, such as the case for matrix-variate Student$-t$ distribution. In
this paper, we propose a unified framework which is used not only to introduce
a novel multivariate Student$-t$ process regression model (MV-TPR) for
multi-output prediction, but also to reformulate the multivariate Gaussian
process regression (MV-GPR) that overcomes some limitations of the existing
methods. Both MV-GPR and MV-TPR have closed-form expressions for the marginal
likelihoods and predictive distributions under this unified framework and thus
can adopt the same optimization approaches as used in the conventional GPR. The
usefulness of the proposed methods is illustrated through several simulated and
real data examples. In particular, we verify empirically that MV-TPR has
superiority for the datasets considered, including air quality prediction and
bike rent prediction. At last, the proposed methods are shown to produce
profitable investment strategies in the stock markets.
| stat |
Bayesian inference for spline-based hidden Markov models | B-spline-based hidden Markov models (HMMs), where the emission densities are
specified as mixtures of normalized B-spline basis functions, offer a more
flexible modelling approach to data than conventional parametric HMMs. We
introduce a fully Bayesian framework for inference in these nonparametric
models where the number of states may be unknown along with other model
parameters. We propose the use of a trans-dimensional Markov chain inference
algorithm to identify a parsimonious knot configuration of the B-splines while
model selection regarding the number of states can be performed within a
parallel sampling framework. The feasibility and efficiency of our proposed
methodology is shown in a simulation study. Its explorative use for real data
is demonstrated for activity acceleration data in animals, i.e.
whitetip-sharks. The flexibility of a Bayesian approach allows us to extend the
modelling framework in a straightforward way and we demonstrate this by
developing a hierarchical conditional HMM to analyse human accelerator activity
data to focus on studying small movements and/or inactivity during sleep.
| stat |
On the connection between orthant probabilities and the first passage
time problem | This article describes a new Monte Carlo method for the evaluation of the
orthant probabilities by sampling first passage times of a non-singular
Gaussian discrete time-series across an absorbing boundary. This procedure
makes use of a simulation of several time-series sample paths, aiming to record
their first crossing instants. Thus, the computation of the orthant
probabilities is traced back to the accurate simulation of a non-singular
Gaussian discrete-time series. Moreover, if the simulation is also efficient,
this method is shown to be more speedy than the others proposed in the
literature. As example, we make use of the Davies-Harte algorithm in the
evaluation of the orthant probabilities associated to the ARFIMA$(0,d,0)$
model. Test results are presented that compare this method with currently
available software.
| stat |
Skeleton Clustering: Dimension-Free Density-based Clustering | We introduce a density-based clustering method called skeleton clustering
that can detect clusters in multivariate and even high-dimensional data with
irregular shapes. To bypass the curse of dimensionality, we propose surrogate
density measures that are less dependent on the dimension but have intuitive
geometric interpretations. The clustering framework constructs a concise
representation of the given data as an intermediate step and can be thought of
as a combination of prototype methods, density-based clustering, and
hierarchical clustering. We show by theoretical analysis and empirical studies
that the skeleton clustering leads to reliable clusters in multivariate and
high-dimensional scenarios.
| stat |
Clustering Left-Censored Multivariate Time-Series | Unsupervised learning seeks to uncover patterns in data. However, different
kinds of noise may impede the discovery of useful substructure from real-world
time-series data. In this work, we focus on mitigating the interference of
left-censorship in the task of clustering. We provide conditions under which
clusters and left-censorship may be identified; motivated by this result, we
develop a deep generative, continuous-time model of time-series data that
clusters while correcting for censorship time. We demonstrate accurate, stable,
and interpretable results on synthetic data that outperform several benchmarks.
To showcase the utility of our framework on real-world problems, we study how
left-censorship can adversely affect the task of disease phenotyping, resulting
in the often incorrect assumption that longitudinal patient data are aligned by
disease stage. In reality, patients at the time of diagnosis are at different
stages of the disease -- both late and early due to differences in when
patients seek medical care and such discrepancy can confound unsupervised
learning algorithms. On two clinical datasets, our model corrects for this form
of censorship and recovers known clinical subtypes.
| stat |
Statistical Inference for Ordinal Predictors in Generalized Linear and
Additive Models with Application to Bronchopulmonary Dysplasia | Discrete but ordered covariates are quite common in applied statistics, and
some regularized fitting procedures have been proposed for proper handling of
ordinal predictors in statistical modeling. In this study, we show how
quadratic penalties on adjacent dummy coefficients of ordinal predictors
proposed in the literature can be incorporated in the framework of generalized
additive models, making tools for statistical inference developed there
available for ordinal predictors as well. Motivated by an application from
neonatal medicine, we discuss whether results obtained when constructing
confidence intervals and testing significance of smooth terms in generalized
additive models are useful with ordinal predictors/penalties as well.
| stat |
A multiple testing framework for diagnostic accuracy studies with
co-primary endpoints | Major advances have been made regarding the utilization of artificial
intelligence in health care. In particular, deep learning approaches have been
successfully applied for automated and assisted disease diagnosis and prognosis
based on complex and high-dimensional data. However, despite all justified
enthusiasm, overoptimistic assessments of predictive performance are still
common. Automated medical testing devices based on machine-learned prediction
models should thus undergo a throughout evaluation before being implemented
into clinical practice. In this work, we propose a multiple testing framework
for (comparative) phase III diagnostic accuracy studies with sensitivity and
specificity as co-primary endpoints. Our approach challenges the frequent
recommendation to strictly separate model selection and evaluation, i.e. to
only assess a single diagnostic model in the evaluation study. We show that our
parametric simultaneous test procedure asymptotically allows strong control of
the family-wise error rate. Moreover, we demonstrate in extensive simulation
studies that our multiple testing strategy on average leads to a better final
diagnostic model and increased statistical power. To plan such studies, we
propose a Bayesian approach to determine the optimal number of models to
evaluate. For this purpose, our algorithm optimizes the expected final model
performance given previous (hold-out) data from the model development phase. We
conclude that an assessment of multiple promising diagnostic models in the same
evaluation study has several advantages when suitable adjustments for multiple
comparisons are implemented.
| stat |
Estimating Conditional Transfer Entropy in Time Series using Mutual
Information and Non-linear Prediction | We propose a new estimator to measure directed dependencies in time series.
The dimensionality of data is first reduced using a new non-uniform embedding
technique, where the variables are ranked according to a weighted sum of the
amount of new information and improvement of the prediction accuracy provided
by the variables. Then, using a greedy approach, the most informative subsets
are selected in an iterative way. The algorithm terminates, when the highest
ranked variable is not able to significantly improve the accuracy of the
prediction as compared to that obtained using the existing selected subsets. In
a simulation study, we compare our estimator to existing state-of-the-art
methods at different data lengths and directed dependencies strengths. It is
demonstrated that the proposed estimator has a significantly higher accuracy
than that of existing methods, especially for the difficult case, where the
data is highly correlated and coupled. Moreover, we show its false detection of
directed dependencies due to instantaneous couplings effect is lower than that
of existing measures. We also show applicability of the proposed estimator on
real intracranial electroencephalography data.
| stat |
SPUX Framework: a Scalable Package for Bayesian Uncertainty
Quantification and Propagation | We present SPUX - a modular framework for Bayesian inference enabling
uncertainty quantification and propagation in linear and nonlinear,
deterministic and stochastic models, and supporting Bayesian model selection.
SPUX can be coupled to any serial or parallel application written in any
programming language, (e.g. including Python, R, Julia, C/C++, Fortran, Java,
or a binary executable), scales effortlessly from serial runs on a personal
computer to parallel high performance computing clusters, and aims to provide a
platform particularly suited to support and foster reproducibility in
computational science. We illustrate SPUX capabilities for a simple yet
representative random walk model, describe how to couple different types of
user applications, and showcase several readily available examples from
environmental sciences. In addition to available state-of-the-art numerical
inference algorithms including EMCEE, PMCMC (PF) and SABC, the open source
nature of the SPUX framework and the explicit description of the hierarchical
parallel SPUX executors should also greatly simplify the implementation and
usage of other inference and optimization techniques.
| stat |
A very short guide to IOI: A general framework for statistical inference
summarised | Integrated organic inference (IOI) is discussed in a concise and informal way
with the aim that the reader is given the gist of what this approach to
statistical inference is about as well as given pointers to further reading.
| stat |
A class of network models recoverable by spectral clustering | Finding communities in networks is a problem that remains difficult, in spite
of the amount of attention it has recently received. The Stochastic Block-Model
(SBM) is a generative model for graphs with "communities" for which, because of
its simplicity, the theoretical understanding has advanced fast in recent
years. In particular, there have been various results showing that simple
versions of spectral clustering using the Normalized Laplacian of the graph can
recover the communities almost perfectly with high probability. Here we show
that essentially the same algorithm used for the SBM and for its extension
called Degree-Corrected SBM, works on a wider class of Block-Models, which we
call Preference Frame Models, with essentially the same guarantees. Moreover,
the parametrization we introduce clearly exhibits the free parameters needed to
specify this class of models, and results in bounds that expose with more
clarity the parameters that control the recovery error in this model class.
| stat |
Spatial hierarchical modeling of threshold exceedances using rate
mixtures | We develop new flexible univariate models for light-tailed and heavy-tailed
data, which extend a hierarchical representation of the generalized Pareto (GP)
limit for threshold exceedances. These models can accommodate departure from
asymptotic threshold stability in finite samples while keeping the asymptotic
GP distribution as a special (or boundary) case and can capture the tails and
the bulk jointly without losing much flexibility. Spatial dependence is modeled
through a latent process, while the data are assumed to be conditionally
independent. Focusing on a gamma-gamma model construction, we design penalized
complexity priors for crucial model parameters, shrinking our proposed spatial
Bayesian hierarchical model toward a simpler reference whose marginal
distributions are GP with moderately heavy tails. Our model can be fitted in
fairly high dimensions using Markov chain Monte Carlo by exploiting the
Metropolis-adjusted Langevin algorithm (MALA), which guarantees fast
convergence of Markov chains with efficient block proposals for the latent
variables. We also develop an adaptive scheme to calibrate the MALA tuning
parameters. Moreover, our model avoids the expensive numerical evaluations of
multifold integrals in censored likelihood expressions. We demonstrate our new
methodology by simulation and application to a dataset of extreme rainfall
events that occurred in Germany. Our fitted gamma-gamma model provides a
satisfactory performance and can be successfully used to predict rainfall
extremes at unobserved locations.
| stat |
A Fast and Greedy Subset-of-Data (SoD) Scheme for Sparsification in
Gaussian processes | In their standard form Gaussian processes (GPs) provide a powerful
non-parametric framework for regression and classificaton tasks. Their one
limiting property is their $\mathcal{O}(N^{3})$ scaling where $N$ is the number
of training data points. In this paper we present a framework for GP training
with sequential selection of training data points using an intuitive selection
metric. The greedy forward selection strategy is devised to target two factors
- regions of high predictive uncertainty and underfit. Under this technique the
complexity of GP training is reduced to $\mathcal{O}(M^{3})$ where $(M \ll N)$
if $M$ data points (out of $N$) are eventually selected. The sequential nature
of the algorithm circumvents the need to invert the covariance matrix of
dimension $N \times N$ and enables the use of favourable matrix inverse update
identities. We outline the algorithm and sequential updates to the posterior
mean and variance. We demonstrate our method on selected one dimensional
functions and show that the loss in accuracy due to using a subset of data
points is marginal compared to the computational gains.
| stat |
Imputing missing values with unsupervised random trees | This work proposes a non-iterative strategy for missing value imputations
which is guided by similarity between observations, but instead of explicitly
determining distances or nearest neighbors, it assigns observations to
overlapping buckets through recursive semi-random hyperplane cuts, in which
weighted averages are determined as imputations for each variable. The quality
of these imputations is oftentimes not as good as that of chained equations,
but the proposed technique is much faster, non-iterative, can make imputations
on new data without re-calculating anything, and scales easily to large and
high-dimensional datasets, providing a significant boost over simple
mean/median imputation in regression and classification metrics with imputed
values when other methods are not feasible.
| stat |
Robust Inference Using the Exponential-Polynomial Divergence | Density-based minimum divergence procedures represent popular techniques in
parametric statistical inference. They combine strong robustness properties
with high (sometimes full) asymptotic efficiency. Among density-based minimum
distance procedures, the methods based on the Bregman-divergence have the
attractive property that the empirical formulation of the divergence does not
require the use of any non-parametric smoothing technique such as kernel
density estimation. The methods based on the density power divergence (DPD)
represent the current standard in this area of research. In this paper, we will
present a more generalized divergence that subsumes the DPD as a special case
and produces several new options providing better compromises between
robustness and efficiency.
| stat |
Design optimization of stochastic complex systems via iterative density
estimation | Reliability-based design optimization (RBDO) provides a rational and sound
framework for finding the optimal design while taking uncertainties into
ac-count. The main issue in implementing RBDO methods, particularly stochastic
simu-lation based ones, is the computational burden arising from the evaluation
of reliability constraints. In this contribution, we propose an efficient
method which ap-proximates the failure probability functions (FPF) to decouple
reliability. Based on the augmentation concept, the approximation of FPF is
equivalent to density estimation of failure design samples. Unlike traditional
density estimation schemes, where the esti-mation is conducted in the entire
design space, in the proposed method we iteratively partition the design space
into several subspaces according to the distribution of fail-ure design
samples. Numerical results of an illustrative example indicate that the
pro-posed method can improve the computational performance considerably.
| stat |
Dirac Delta Regression: Conditional Density Estimation with Clinical
Trials | Personalized medicine seeks to identify the causal effect of treatment for a
particular patient as opposed to a clinical population at large. Most
investigators estimate such personalized treatment effects by regressing the
outcome of a randomized clinical trial (RCT) on patient covariates. The
realized value of the outcome may however lie far from the conditional
expectation. We therefore introduce a method called Dirac Delta Regression
(DDR) that estimates the entire conditional density from RCT data in order to
visualize the probabilities across all possible treatment outcomes. DDR
transforms the outcome into a set of asymptotically Dirac delta distributions
and then estimates the density using non-linear regression. The algorithm can
identify significant patient-specific treatment effects even when no population
level effect exists. Moreover, DDR outperforms state-of-the-art algorithms in
conditional density estimation on average regardless of the need for causal
inference.
| stat |
A Multi-Phase Approach for Product Hierarchy Forecasting in Supply Chain
Management: Application to MonarchFx Inc | Hierarchical time series demands exist in many industries and are often
associated with the product, time frame, or geographic aggregations.
Traditionally, these hierarchies have been forecasted using top-down,
bottom-up, or middle-out approaches. The question we aim to answer is how to
utilize child-level forecasts to improve parent-level forecasts in a
hierarchical supply chain. Improved forecasts can be used to considerably
reduce logistics costs, especially in e-commerce. We propose a novel
multi-phase hierarchical (MPH) approach. Our method involves forecasting each
series in the hierarchy independently using machine learning models, then
combining all forecasts to allow a second phase model estimation at the parent
level. Sales data from MonarchFx Inc. (a logistics solutions provider) is used
to evaluate our approach and compare it to bottom-up and top-down methods. Our
results demonstrate an 82-90% improvement in forecast accuracy using the
proposed approach. Using the proposed method, supply chain planners can derive
more accurate forecasting models to exploit the benefit of multivariate data.
| stat |
Fair Data Adaptation with Quantile Preservation | Fairness of classification and regression has received much attention
recently and various, partially non-compatible, criteria have been proposed.
The fairness criteria can be enforced for a given classifier or, alternatively,
the data can be adapated to ensure that every classifier trained on the data
will adhere to desired fairness criteria. We present a practical data adaption
method based on quantile preservation in causal structural equation models. The
data adaptation is based on a presumed counterfactual model for the data. While
the counterfactual model itself cannot be verified experimentally, we show that
certain population notions of fairness are still guaranteed even if the
counterfactual model is misspecified. The precise nature of the fulfilled
non-causal fairness notion (such as demographic parity, separation or
sufficiency) depends on the structure of the underlying causal model and the
choice of resolving variables. We describe an implementation of the proposed
data adaptation procedure based on Random Forests and demonstrate its practical
use on simulated and real-world data.
| stat |
Automated Augmented Conjugate Inference for Non-conjugate Gaussian
Process Models | We propose automated augmented conjugate inference, a new inference method
for non-conjugate Gaussian processes (GP) models. Our method automatically
constructs an auxiliary variable augmentation that renders the GP model
conditionally conjugate. Building on the conjugate structure of the augmented
model, we develop two inference methods. First, a fast and scalable stochastic
variational inference method that uses efficient block coordinate ascent
updates, which are computed in closed form. Second, an asymptotically correct
Gibbs sampler that is useful for small datasets. Our experiments show that our
method are up two orders of magnitude faster and more robust than existing
state-of-the-art black-box methods.
| stat |
On the geometry of Stein variational gradient descent | Bayesian inference problems require sampling or approximating
high-dimensional probability distributions. The focus of this paper is on the
recently introduced Stein variational gradient descent methodology, a class of
algorithms that rely on iterated steepest descent steps with respect to a
reproducing kernel Hilbert space norm. This construction leads to interacting
particle systems, the mean-field limit of which is a gradient flow on the space
of probability distributions equipped with a certain geometrical structure. We
leverage this viewpoint to shed some light on the convergence properties of the
algorithm, in particular addressing the problem of choosing a suitable positive
definite kernel function. Our analysis leads us to considering certain
nondifferentiable kernels with adjusted tails. We demonstrate significant
performs gains of these in various numerical experiments.
| stat |
Statistical analysis of the water level of Huang He river (Yellow river)
in China | Very high water levels of the large rivers are extremely dangerous events
that can lead to large floods and loss of property and thousands and even tens
of thousands human lives. The information from the systematical monitoring of
the water levels allows us to obtain probability distributions for the
extremely high values of the water levels of the rivers of interest. In this
article we study time series containing information from more than 10 years of
satellite observation of the water level of the Huang He river (Yellow river)
in China. We show that the corresponding extreme values distribution is the
Weibull distribution and determine the parameters of the distribution. The
obtained results may help for evaluation of risks associated with floods for
the population and villages in the corresponding region of the Huang He river.
| stat |
Bayesian Multinomial Logistic Normal Models through Marginally Latent
Matrix-T Processes | Bayesian multinomial logistic-normal (MLN) models are popular for the
analysis of sequence count data (e.g., microbiome or gene expression data) due
to their ability to model multivariate count data with complex covariance
structure. However, existing implementations of MLN models are limited to
handling small data sets due to the non-conjugacy of the multinomial and
logistic-normal distributions. We introduce MLN models which can be written as
marginally latent matrix-t process (LTP) models. Marginally LTP models describe
a flexible class of generalized linear regression, non-linear regression, and
time series models. We develop inference schemes for Marginally LTP models and,
through application to MLN models, demonstrate that our inference schemes are
both highly accurate and often 4-5 orders of magnitude faster than MCMC.
| stat |
Sharp threshold for alignment of graph databases with Gaussian weights | We study the fundamental limits for reconstruction in weighted graph (or
matrix) database alignment. We consider a model of two graphs where $\pi^*$ is
a planted uniform permutation and all pairs of edge weights $(A_{i,j},
B_{\pi^*(i),\pi^*(j)})_{1 \leq i<j \leq n}$ are i.i.d. pairs of Gaussian
variables with zero mean, unit variance and correlation parameter $\rho \in
[0,1]$. We prove that there is a sharp threshold for exact recovery of $\pi^*$:
if $n \rho^2 \geq (4+\varepsilon) \log n + \omega(1)$ for some $\varepsilon>0$,
there is an estimator $\hat{\pi}$ -- namely the MAP estimator -- based on the
observation of databases $A,B$ that achieves exact reconstruction with high
probability. Conversely, if $n \rho^2 \leq 4 \log n - \log \log n - \omega(1)$,
then any estimator $\hat{\pi}$ verifies $\hat{\pi}=\pi$ with probability
$o(1)$. This result shows that the information-theoretic threshold for exact
recovery is the same as the one obtained for detection in a recent work by Y.
Wu, J. Xu and S. Yu: in other words, for Gaussian weighted graph alignment, the
problem of reconstruction is not more difficult than that of detection. Though
the reconstruction task was already well understood for vector-shaped database
alignment (that is taking signal of the form $(u_i, v_{\pi^*(i)})_{1 \leq i\leq
n}$ where $(u_i, v_{\pi^*(i)})$ are i.i.d. pairs in $\mathbb{R}^{d_u} \times
\mathbb{R}^{d_v}$), its formulation for graph (or matrix) databases brings a
drastically different problem for which the hard phase is conjectured to be
huge. The study is based on the analysis of the MAP estimator, and proofs rely
on proper use of the correlation structure of energies of permutations.
| stat |
Learning Models with Uniform Performance via Distributionally Robust
Optimization | A common goal in statistics and machine learning is to learn models that can
perform well against distributional shifts, such as latent heterogeneous
subpopulations, unknown covariate shifts, or unmodeled temporal effects. We
develop and analyze a distributionally robust stochastic optimization (DRO)
framework that learns a model providing good performance against perturbations
to the data-generating distribution. We give a convex formulation for the
problem, providing several convergence guarantees. We prove finite-sample
minimax upper and lower bounds, showing that distributional robustness
sometimes comes at a cost in convergence rates. We give limit theorems for the
learned parameters, where we fully specify the limiting distribution so that
confidence intervals can be computed. On real tasks including generalizing to
unknown subpopulations, fine-grained recognition, and providing good tail
performance, the distributionally robust approach often exhibits improved
performance.
| stat |
Solving Bayesian Inverse Problems via Variational Autoencoders | In recent years, the field of machine learning has made phenomenal progress
in the pursuit of simulating real-world data generation processes. One notable
example of such success is the variational autoencoder (VAE). In this work,
with a small shift in perspective, we leverage and adapt VAEs for a different
purpose: uncertainty quantification in scientific inverse problems. We
introduce UQ-VAE: a flexible, adaptive, hybrid data/model-informed framework
for training neural networks capable of rapid modelling of the posterior
distribution representing the unknown parameter of interest. Specifically, from
divergence-based variational inference, our framework is derived such that most
of the information usually present in scientific inverse problems is fully
utilized in the training procedure. Additionally, this framework includes an
adjustable hyperparameter that allows selection of the notion of distance
between the posterior model and the target distribution. This introduces more
flexibility in controlling how optimization directs the learning of the
posterior model. Further, this framework possesses an inherent adaptive
optimization property that emerges through the learning of the posterior
uncertainty.
| stat |
Parallel Predictive Entropy Search for Multi-objective Bayesian
Optimization with Constraints | Real-world problems often involve the optimization of several objectives
under multiple constraints. Furthermore, we may not have an expression for each
objective or constraint; they may be expensive to evaluate; and the evaluations
can be noisy. These functions are referred to as black-boxes. Bayesian
optimization (BO) can efficiently solve the problems described. For this, BO
iteratively fits a model to the observations of each black-box. The models are
then used to choose where to evaluate the black-boxes next, with the goal of
solving the optimization problem in a few iterations. In particular, they guide
the search for the problem solution, and avoid evaluations in regions of little
expected utility. A limitation, however, is that current BO methods for these
problems choose a point at a time at which to evaluate the black-boxes. If the
expensive evaluations can be carried out in parallel (as when a cluster of
computers is available), this results in a waste of resources. Here, we
introduce PPESMOC, Parallel Predictive Entropy Search for Multi-objective
Optimization with Constraints, a BO strategy for solving the problems
described. PPESMOC selects, at each iteration, a batch of input locations at
which to evaluate the black-boxes, in parallel, to maximally reduce the entropy
of the problem solution. To our knowledge, this is the first batch method for
constrained multi-objective BO. We present empirical evidence in the form of
synthetic, benchmark and real-world experiments that illustrate the
effectiveness of PPESMOC.
| stat |
Meta-learning One-class Classifiers with Eigenvalue Solvers for
Supervised Anomaly Detection | Neural network-based anomaly detection methods have shown to achieve high
performance. However, they require a large amount of training data for each
task. We propose a neural network-based meta-learning method for supervised
anomaly detection. The proposed method improves the anomaly detection
performance on unseen tasks, which contains a few labeled normal and anomalous
instances, by meta-training with various datasets. With a meta-learning
framework, quick adaptation to each task and its effective backpropagation are
important since the model is trained by the adaptation for each epoch. Our
model enables them by formulating adaptation as a generalized eigenvalue
problem with one-class classification; its global optimum solution is obtained,
and the solver is differentiable. We experimentally demonstrate that the
proposed method achieves better performance than existing anomaly detection and
few-shot learning methods on various datasets.
| stat |
Estimation of Classification Rules from Partially Classified Data | We consider the situation where the observed sample contains some
observations whose class of origin is known (that is, they are classified with
respect to the g underlying classes of interest), and where the remaining
observations in the sample are unclassified (that is, their class labels are
unknown). For class-conditional distributions taken to be known up to a vector
of unknown parameters, the aim is to estimate the Bayes' rule of allocation for
the allocation of subsequent unclassified observations. Estimation on the basis
of both the classified and unclassified data can be undertaken in a
straightforward manner by fitting a g-component mixture model by maximum
likelihood (ML) via the EM algorithm in the situation where the observed data
can be assumed to be an observed random sample from the adopted mixture
distribution. This assumption applies if the missing-data mechanism is
ignorable in the terminology pioneered by Rubin (1976). An initial likelihood
approach was to use the so-called classification ML approach whereby the
missing labels are taken to be parameters to be estimated along with the
parameters of the class-conditional distributions. However, as it can lead to
inconsistent estimates, the focus of attention switched to the mixture ML
approach after the appearance of the EM algorithm (Dempster et al., 1977).
Particular attention is given here to the asymptotic relative efficiency (ARE)
of the Bayes' rule estimated from a partially classified sample. Lastly, we
consider briefly some recent results in situations where the missing label
pattern is non-ignorable for the purposes of ML estimation for the mixture
model.
| stat |
A Case Study of Nonresponse Bias Analysis | Nonresponse bias is a widely prevalent problem for data collections. We
develop a ten-step exemplar to guide nonresponse bias analysis (NRBA) in
cross-sectional studies, and apply these steps to the Early Childhood
Longitudinal Study, Kindergarten Class of 2010-11. A key step is the
construction of indices of nonresponse bias based on proxy pattern-mixture
models for survey variables of interest. A novel feature is to characterize the
strength of evidence about nonresponse bias contained in these indices, based
on the strength of relationship of the characteristics in the nonresponse
adjustment with the key survey variables. Our NRBA incorporates missing at
random and missing not at random mechanisms, and all analyses can be done
straightforwardly with standard statistical software.
| stat |
Multiple imputation in functional regression with applications to EEG
data in a depression study | Current source density (CSD) power asymmetry, a measure derived from
electroencephalography (EEG), is a potential biomarker for major depressive
disorder (MDD). Though this measure is functional in nature (defined on the
frequency domain), it is typically reduced to a scalar value prior to analysis,
possibly obscuring the relationship between brain function and MDD. To overcome
this issue, we sought to fit a functional regression model to estimate the
association between CSD power asymmetry and MDD diagnostic status, adjusting
for age, sex, cognitive ability, and handedness using data from a large
clinical study. Unfortunately, nearly 40\% of the observations were missing
either their functional EEG data, their cognitive ability score, or both. In
order to take advantage of all of the available data, we propose an extension
to multiple imputation by chained equations that handles both scalar and
functional data. We also propose an extension to Rubin's Rules for pooling
estimates from the multiply imputed data sets in order to conduct valid
inference. We investigate the performance of the proposed extensions in a
simulation study and apply them to our clinical study data. Our analysis
reveals that the association between CSD power asymmetry and diagnostic status
depends on both age and sex.
| stat |
Bayesian Computing in the Undergraduate Statistics Curriculum | Bayesian statistics has gained great momentum since the computational
developments of the 1990s. Gradually, advances in Bayesian methodology and
software have made Bayesian techniques much more accessible to applied
statisticians and, in turn, have potentially transformed Bayesian education at
the undergraduate level. This article provides an overview on the various
options for implementing Bayesian computational methods motivated to achieve
particular learning outcomes. The advantages and disadvantages of each
computational method are described based on the authors' experience in using
these methods in the classroom. The goal is to present guidance on the choice
of computation for the instructors who are introducing Bayesian methods in
their undergraduate statistics curriculum.
| stat |
Sparse Regression with Multi-type Regularized Feature Modeling | Within the statistical and machine learning literature, regularization
techniques are often used to construct sparse (predictive) models. Most
regularization strategies only work for data where all predictors are treated
identically, such as Lasso regression for (continuous) predictors treated as
linear effects. However, many predictive problems involve different types of
predictors and require a tailored regularization term. We propose a multi-type
Lasso penalty that acts on the objective function as a sum of subpenalties, one
for each type of predictor. As such, we allow for predictor selection and level
fusion within a predictor in a data-driven way, simultaneous with the parameter
estimation process. We develop a new estimation strategy for convex predictive
models with this multi-type penalty. Using the theory of proximal operators,
our estimation procedure is computationally efficient, partitioning the overall
optimization problem into easier to solve subproblems, specific for each
predictor type and its associated penalty. Earlier research applies
approximations to non-differentiable penalties to solve the optimization
problem. The proposed SMuRF algorithm removes the need for approximations and
achieves a higher accuracy and computational efficiency. This is demonstrated
with an extensive simulation study and the analysis of a case-study on
insurance pricing analytics.
| stat |
Positive-Unlabelled Survival Data Analysis | In this paper, we consider a novel framework of positive-unlabeled data in
which as positive data survival times are observed for subjects who have events
during the observation time as positive data and as unlabeled data censoring
times are observed but whether the event occurs or not are unknown for some
subjects. We consider two cases: (1) when censoring time is observed in
positive data, and (2) when it is not observed. For both cases, we developed
parametric models, nonparametric models, and machine learning models and the
estimation strategies for these models. Simulation studies show that under this
data setup, traditional survival analysis may yield severely biased results,
while the proposed estimation method can provide valid results.
| stat |
ABCpy: A High-Performance Computing Perspective to Approximate Bayesian
Computation | ABCpy is a highly modular scientific library for Approximate Bayesian
Computation (ABC) written in Python. The main contribution of this paper is to
document a software engineering effort that enables domain scientists to easily
apply ABC to their research without being ABC experts; using ABCpy they can
easily run large parallel simulations without much knowledge about
parallelization. Further, ABCpy enables ABC experts to easily develop new
inference schemes and evaluate them in a standardized environment and to extend
the library with new algorithms. These benefits come mainly from the modularity
of ABCpy. We give an overview of the design of ABCpy and provide a performance
evaluation concentrating on parallelization. This points us towards the
inherent imbalance in some of the ABC algorithms. We develop a dynamic
scheduling MPI implementation to mitigate this issue and evaluate the various
ABC algorithms according to their adaptability towards high-performance
computing.
| stat |
Spline Analysis of Biomarker Data Pooled From Multiple Matched/Nested
Case-Control Studies | Pooling biomarker data across multiple studies enables researchers to get
more precise estimates of the association between biomarker exposure
measurements and disease risks due to increased sample sizes. However,
biomarker measurements vary significantly across different assays and
laboratories, and therefore calibration of the local laboratory measurements to
a reference laboratory is necessary before pooling data. We propose two methods
that can estimate a nonlinear relationship between biomarker exposure
measurements and disease risks using spline functions with a nested
case-control study design: full calibration and internalized calibration. The
full calibration method calibrates all observations using a study-specific
calibration model while the internalized calibration method only calibrates
observations that do not have reference laboratory measurements available. We
compare the two methods with a naive method whereby data are pooled without
calibration. We find that: (1) Internalized and full calibration methods have
substantially better performance than the naive method in terms of average
relative bias and coverage rate. (2) Full calibration is more robust than
internalized calibration when the size of calibration subsets varies. We apply
our methods to a pooling project with nested case-control study design to
estimate the association of circulating Vitamin D levels with the risk of
colorectal cancer.
| stat |
Asymptotic errors for convex penalized linear regression beyond Gaussian
matrices | We consider the problem of learning a coefficient vector $x_{0}$ in $R^{N}$
from noisy linear observations $y=Fx_{0}+w$ in $R^{M}$ in the high dimensional
limit $M,N$ to infinity with $\alpha=M/N$ fixed. We provide a rigorous
derivation of an explicit formula -- first conjectured using heuristic methods
from statistical physics -- for the asymptotic mean squared error obtained by
penalized convex regression estimators such as the LASSO or the elastic net,
for a class of very generic random matrices corresponding to rotationally
invariant data matrices with arbitrary spectrum. The proof is based on a
convergence analysis of an oracle version of vector approximate message-passing
(oracle-VAMP) and on the properties of its state evolution equations. Our
method leverages on and highlights the link between vector approximate
message-passing, Douglas-Rachford splitting and proximal descent algorithms,
extending previous results obtained with i.i.d. matrices for a large class of
problems. We illustrate our results on some concrete examples and show that
even though they are asymptotic, our predictions agree remarkably well with
numerics even for very moderate sizes.
| stat |
A Bayesian Choice Model for Eliminating Feedback Loops | Self-reinforcing feedback loops in personalization systems are typically
caused by users choosing from a limited set of alternatives presented
systematically based on previous choices. We propose a Bayesian choice model
built on Luce axioms that explicitly accounts for users' limited exposure to
alternatives. Our model is fair---it does not impose negative bias towards
unpresented alternatives, and practical---preference estimates are accurately
inferred upon observing a small number of interactions. It also allows
efficient sampling, leading to a straightforward online presentation mechanism
based on Thompson sampling. Our approach achieves low regret in learning to
present upon exploration of only a small fraction of possible presentations.
The proposed structure can be reused as a building block in interactive
systems, e.g., recommender systems, free of feedback loops.
| stat |
An Evaluation of Change Point Detection Algorithms | Change point detection is an important part of time series analysis, as the
presence of a change point indicates an abrupt and significant change in the
data generating process. While many algorithms for change point detection
exist, little attention has been paid to evaluating their performance on
real-world time series. Algorithms are typically evaluated on simulated data
and a small number of commonly-used series with unreliable ground truth.
Clearly this does not provide sufficient insight into the comparative
performance of these algorithms. Therefore, instead of developing yet another
change point detection method, we consider it vastly more important to properly
evaluate existing algorithms on real-world data. To achieve this, we present
the first data set specifically designed for the evaluation of change point
detection algorithms, consisting of 37 time series from various domains. Each
time series was annotated by five expert human annotators to provide ground
truth on the presence and location of change points. We analyze the consistency
of the human annotators, and describe evaluation metrics that can be used to
measure algorithm performance in the presence of multiple ground truth
annotations. Subsequently, we present a benchmark study where 14 existing
algorithms are evaluated on each of the time series in the data set. This study
shows that binary segmentation (Scott and Knott, 1974) and Bayesian online
change point detection (Adams and MacKay, 2007) are among the best performing
methods. Our aim is that this data set will serve as a proving ground in the
development of novel change point detection algorithms.
| stat |
Caveats for information bottleneck in deterministic scenarios | Information bottleneck (IB) is a method for extracting information from one
random variable $X$ that is relevant for predicting another random variable
$Y$. To do so, IB identifies an intermediate "bottleneck" variable $T$ that has
low mutual information $I(X;T)$ and high mutual information $I(Y;T)$. The "IB
curve" characterizes the set of bottleneck variables that achieve maximal
$I(Y;T)$ for a given $I(X;T)$, and is typically explored by maximizing the "IB
Lagrangian", $I(Y;T) - \beta I(X;T)$. In some cases, $Y$ is a deterministic
function of $X$, including many classification problems in supervised learning
where the output class $Y$ is a deterministic function of the input $X$. We
demonstrate three caveats when using IB in any situation where $Y$ is a
deterministic function of $X$: (1) the IB curve cannot be recovered by
maximizing the IB Lagrangian for different values of $\beta$; (2) there are
"uninteresting" trivial solutions at all points of the IB curve; and (3) for
multi-layer classifiers that achieve low prediction error, different layers
cannot exhibit a strict trade-off between compression and prediction, contrary
to a recent proposal. We also show that when $Y$ is a small perturbation away
from being a deterministic function of $X$, these three caveats arise in an
approximate way. To address problem (1), we propose a functional that, unlike
the IB Lagrangian, can recover the IB curve in all cases. We demonstrate the
three caveats on the MNIST dataset.
| stat |
Semi-Generative Modelling: Covariate-Shift Adaptation with Cause and
Effect Features | Current methods for covariate-shift adaptation use unlabelled data to compute
importance weights or domain-invariant features, while the final model is
trained on labelled data only. Here, we consider a particular case of covariate
shift which allows us also to learn from unlabelled data, that is, combining
adaptation with semi-supervised learning. Using ideas from causality, we argue
that this requires learning with both causes, $X_C$, and effects, $X_E$, of a
target variable, $Y$, and show how this setting leads to what we call a
semi-generative model, $P(Y,X_E|X_C,\theta)$. Our approach is robust to domain
shifts in the distribution of causal features and leverages unlabelled data by
learning a direct map from causes to effects. Experiments on synthetic data
demonstrate significant improvements in classification over purely-supervised
and importance-weighting baselines.
| stat |
Adaptive Change Point Monitoring for High-Dimensional Data | In this paper, we propose a class of monitoring statistics for a mean shift
in a sequence of high-dimensional observations. Inspired by the recent
U-statistic based retrospective tests developed by Wang et al.(2019) and Zhang
et al.(2020), we advance the U-statistic based approach to the sequential
monitoring problem by developing a new adaptive monitoring procedure that can
detect both dense and sparse changes in real-time. Unlike Wang et al.(2019) and
Zhang et al.(2020), where self-normalization was used in their tests, we
instead introduce a class of estimators for $q$-norm of the covariance matrix
and prove their ratio consistency. To facilitate fast computation, we further
develop recursive algorithms to improve the computational efficiency of the
monitoring procedure. The advantage of the proposed methodology is demonstrated
via simulation studies and real data illustrations.
| stat |
A Modified Sequential Probability Ratio Test | We describe a modified sequential probability ratio test that can be used to
reduce the average sample size required to perform statistical hypothesis tests
at specified levels of significance and power. Examples are provided for $z$
tests, $t$ tests, and tests of binomial success probabilities. A description of
a software package to implement the test designs is provided. We compare the
sample sizes required in fixed design tests conducted at 5$\%$ significance
levels to the average sample sizes required in sequential tests conducted at
0.5$\%$ significance levels, and we find that the two sample sizes are
approximately equal.
| stat |
Learning Joint Nonlinear Effects from Single-variable Interventions in
the Presence of Hidden Confounders | We propose an approach to estimate the effect of multiple simultaneous
interventions in the presence of hidden confounders. To overcome the problem of
hidden confounding, we consider the setting where we have access to not only
the observational data but also sets of single-variable interventions in which
each of the treatment variables is intervened on separately. We prove
identifiability under the assumption that the data is generated from a
nonlinear continuous structural causal model with additive Gaussian noise. In
addition, we propose a simple parameter estimation method by pooling all the
data from different regimes and jointly maximizing the combined likelihood. We
also conduct comprehensive experiments to verify the identifiability result as
well as to compare the performance of our approach against a baseline on both
synthetic and real-world data.
| stat |
New weighted $L^2$-type tests for the inverse Gaussian distribution | We propose a new class of goodness-of-fit tests for the inverse Gaussian
distribution. The proposed tests are weighted $L^2$-type tests depending on a
tuning parameter. We develop the asymptotic theory under the null hypothesis
and under a broad class of alternative distributions. These results are used to
show that the parametric bootstrap procedure, which we employ to implement the
test, is asymptotically valid and that the whole test procedure is consistent.
A comparative simulation study for finite sample sizes shows that the new
procedure is competitive to classical and recent tests, outperforming these
other methods almost uniformly over a large set of alternative distributions.
The use of the newly proposed test is illustrated with two observed data sets.
| stat |
Embedded Bayesian Network Classifiers | Low-dimensional probability models for local distribution functions in a
Bayesian network include decision trees, decision graphs, and causal
independence models. We describe a new probability model for discrete Bayesian
networks, which we call an embedded Bayesian network classifier or EBNC. The
model for a node $Y$ given parents $\bf X$ is obtained from a (usually
different) Bayesian network for $Y$ and $\bf X$ in which $\bf X$ need not be
the parents of $Y$. We show that an EBNC is a special case of a softmax
polynomial regression model. Also, we show how to identify a non-redundant set
of parameters for an EBNC, and describe an asymptotic approximation for
learning the structure of Bayesian networks that contain EBNCs. Unlike the
decision tree, decision graph, and causal independence models, we are unaware
of a semantic justification for the use of these models. Experiments are needed
to determine whether the models presented in this paper are useful in practice.
| stat |
A psychometric modeling approach to fuzzy rating data | Modeling fuzziness and imprecision in human rating data is a crucial problem
in many research areas, including applied statistics, behavioral, social, and
health sciences. Because of the interplay between cognitive, affective, and
contextual factors, the process of answering survey questions is a complex
task, which can barely be captured by standard (crisp) rating responses. Fuzzy
rating scales have progressively been adopted to overcome some of the
limitations of standard rating scales, including their inability to disentangle
decision uncertainty from individual responses. The aim of this article is to
provide a novel fuzzy scaling procedure which uses Item Response Theory trees
(IRTrees) as a psychometric model for the stage-wise latent response process.
In so doing, fuzziness of rating data is modeled using the overall rater's
pattern of responses instead of being computed using a single-item based
approach. This offers a consistent system for interpreting fuzziness in terms
of individual-based decision uncertainty. A simulation study and two empirical
applications are adopted to assess the characteristics of the proposed model
and provide converging results about its effectiveness in modeling fuzziness
and imprecision in rating data.
| stat |
Is suicide mortality associated with neighbourhood social fragmentation
and deprivation? A Dutch register-based case-control study using
individualized neighbourhoods | Background: Neighbourhood social fragmentation and socioeconomic deprivation
seem to be associated with suicide mortality. However, results are
inconclusive, which might be because dynamics in the social context are not
well-represented by administratively bounded neighbourhoods at baseline. We
used individualized neighbourhoods to examine associations between suicide
mortality, social fragmentation, and deprivation for the total population as
well as by sex and age group. Methods: Using a nested case-control design, all
suicides aged 18-64 years between 2007 and 2016 were selected from longitudinal
Dutch register data and matched with 10 random controls. Indices for social
fragmentation and deprivation were calculated annually for 300, 600, and 1,000
metre circular buffers around each subject's residential address. Results:
Suicide mortality was significantly higher in neighbourhoods with high
deprivation and social fragmentation. Accounting for individual characteristics
largely attenuated these associations. Suicide mortality remained significantly
higher for women living in highly fragmented neighbourhoods in the fully
adjusted model. Age-stratified analyses indicate associations with
neighbourhood fragmentation among women in older age groups (40-64 years) only.
Among men, suicide risk was lower in fragmented neighbourhoods for
18-39-year-olds and for short-term residents. In deprived neighbourhoods, the
suicide risk was lower for 40-64-year-old men and long-term residents.
Associations between neighbourhood characteristics and suicide mortality were
comparable across buffer sizes. Conclusion: Our findings suggest that next to
individual characteristics, the social and economic context within which people
live may both enhance and buffer the risk of suicide.
| stat |
Assessment of the influence of features on a classification problem: an
application to COVID-19 patients | This paper deals with an important subject in classification problems
addressed by machine learning techniques: the evaluation of the influence of
each of the features on the classification of individuals. Specifically, a
measure of that influence is introduced using the Shapley value of cooperative
games. In addition, an axiomatic characterisation of the proposed measure is
provided based on properties of efficiency and balanced contributions.
Furthermore, some experiments have been designed in order to validate the
appropriate performance of such measure. Finally, the methodology introduced is
applied to a sample of COVID-19 patients to study the influence of certain
demographic or risk factors on various events of interest related to the
evolution of the disease.
| stat |
Time series copula models using d-vines and v-transforms | An approach to modelling volatile financial return series using stationary
d-vine copula processes combined with Lebesgue-measure-preserving
transformations known as v-transforms is proposed. By developing a method of
stochastically inverting v-transforms, models are constructed that can describe
both stochastic volatility in the magnitude of price movements and serial
correlation in their directions. In combination with parametric marginal
distributions it is shown that these models can rival and sometimes outperform
well-known models in the extended GARCH family.
| stat |
intRinsic: an R package for model-based estimation of the intrinsic
dimension of a dataset | The estimation of the intrinsic dimension of a dataset is a fundamental step
in most dimensionality reduction techniques. This article illustrates
intRinsic, an R package that implements novel state-of-the-art likelihood-based
estimators of the intrinsic dimension of a dataset. In detail, the methods
included in this package are the TWO-NN, Gride, and Hidalgo models. To allow
these novel estimators to be easily accessible, the package contains a few
high-level, intuitive functions that rely on a broader set of efficient,
low-level routines. intRinsic encompasses models that fall into two categories:
homogeneous and heterogeneous intrinsic dimension estimators. The first
category contains the TWO-NN and Gride models. The functions dedicated to these
two methods carry out inference under both the frequentist and Bayesian
frameworks. In the second category we find Hidalgo, a Bayesian mixture model,
for which an efficient Gibbs sampler is implemented. After discussing the
theoretical background, we demonstrate the performance of the models on
simulated datasets. This way, we can assess the results by comparing them with
the ground truth. Then, we employ the package to study the intrinsic dimension
of the Alon dataset, obtained from a famous microarray experiment. We show how
the estimation of homogeneous and heterogeneous intrinsic dimensions allows us
to gain valuable insights about the topological structure of a dataset.
| stat |
Modified Jaccard Index Analysis and Adaptive Feature Selection for
Location Fingerprinting with Limited Computational Complexity | We propose an approach for fingerprinting-based positioning which reduces the
data requirements and computational complexity of the online positioning stage.
It is based on a segmentation of the entire region of interest into subregions,
identification of candidate subregions during the online-stage, and position
estimation using a preselected subset of relevant features. The subregion
selection uses a modified Jaccard index which quantifies the similarity between
the features observed by the user and those available within the reference
fingerprint map. The adaptive feature selection is achieved using an adaptive
forward-backward greedy search which determines a subset of features for each
subregion, relevant with respect to a given fingerprinting-based positioning
method. In an empirical study using signals of opportunity for fingerprinting
the proposed subregion and feature selection reduce the processing time during
the online-stage by a factor of about 10 while the positioning accuracy does
not deteriorate significantly. In fact, in one of the two study cases the 90th
percentile of the circular error increased by 7.5% while in the other study
case we even found a reduction of the corresponding circular error by 30%.
| stat |
Design- and Model-Based Approaches to Small-Area Estimation in a Low and
Middle Income Country Context: Comparisons and Recommendations | The need for rigorous and timely health and demographic summaries has
provided the impetus for an explosion in geographic studies, with a common
approach being the production of pixel-level maps, particularly in low and
middle income countries. In this context, household surveys are a major source
of data, usually with a two-stage cluster design with stratification by region
and urbanicity. Accurate estimates are of crucial interest for precision public
health policy interventions, but many current studies take a cavalier approach
to acknowledging the sampling design, while presenting results at a fine
geographic scale. In this paper we investigate the extent to which accounting
for sample design can affect predictions at the aggregate level, which is
usually the target of inference. We describe a simulation study in which
realistic sampling frames are created for Kenya, based on population and
demographic information, with a survey design that mimics a Demographic Health
Survey (DHS). We compare the predictive performance of various commonly-used
models. We also describe a cluster level model with a discrete spatial
smoothing prior that has not been previously used, but provides reliable
inference. We find that including stratification and cluster level random
effects can improve predictive performance. Spatially smoothed direct
(weighted) estimates were robust to priors and survey design. Continuous
spatial models performed well in the presence of fine scale variation; however,
these models require the most "hand holding". Subsequently, we examine how the
models perform on real data; specifically we model the prevalence of secondary
education for women aged 20-29 using data from the 2014 Kenya DHS.
| stat |
Emotion Transfer Using Vector-Valued Infinite Task Learning | Style transfer is a significant problem of machine learning with numerous
successful applications. In this work, we present a novel style transfer
framework building upon infinite task learning and vector-valued reproducing
kernel Hilbert spaces. We instantiate the idea in emotion transfer where the
goal is to transform facial images to different target emotions. The proposed
approach provides a principled way to gain explicit control over the continuous
style space. We demonstrate the efficiency of the technique on popular facial
emotion benchmarks, achieving low reconstruction cost and high emotion
classification accuracy.
| stat |
On the Detectability of Conflict: a Remote Sensing Study of the Rohingya
Conflict | The detection and quantification of conflict through remote sensing
modalities represents a challenging but crucial aspect of human rights
monitoring. In this work we demonstrate how utilizing multi-modal data sources
can help build a comprehensive picture of conflict and human displacement,
using the Rohingya conflict in the state of Rakhine, Myanmar as a case study.
We show that time series analysis of fire detections from the Moderate
Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging
Radiometer Suite (VIIRS) can reveal anomalous spatial and temporal
distributions of fires related to conflict. This work also shows that Synthetic
Aperture Radar (SAR) backscatter and coherence data can detect the razing and
burning of buildings and villages, even in cloudy conditions. These techniques
may be further developed in the future to enable the monitoring and detection
of signals originating from these types of conflict.
| stat |
Filtering the rejection set while preserving false discovery rate
control | Scientific hypotheses in a variety of applications have domain-specific
structures, such as the tree structure of the International Classification of
Diseases (ICD), the directed acyclic graph structure of the Gene Ontology (GO),
or the spatial structure in genome-wide association studies. In the context of
multiple testing, the resulting relationships among hypotheses can create
redundancies among rejections that hinder interpretability. This leads to the
practice of filtering rejection sets obtained from multiple testing procedures,
which may in turn invalidate their inferential guarantees. We propose Focused
BH, a simple, flexible, and principled methodology to adjust for the
application of any pre-specified filter. We prove that Focused BH controls the
false discovery rate under various conditions, including when the filter
satisfies an intuitive monotonicity property and the p-values are positively
dependent. We demonstrate in simulations that Focused BH performs well across a
variety of settings, and illustrate this method's practical utility via
analyses of real datasets based on ICD and GO.
| stat |
A study of data and label shift in the LIME framework | LIME is a popular approach for explaining a black-box prediction through an
interpretable model that is trained on instances in the vicinity of the
predicted instance. To generate these instances, LIME randomly selects a subset
of the non-zero features of the predicted instance. After that, the perturbed
instances are fed into the black-box model to obtain labels for these, which
are then used for training the interpretable model. In this study, we present a
systematic evaluation of the interpretable models that are output by LIME on
the two use-cases that were considered in the original paper introducing the
approach; text classification and object detection. The investigation shows
that the perturbation and labeling phases result in both data and label shift.
In addition, we study the correlation between the shift and the fidelity of the
interpretable model and show that in certain cases the shift negatively
correlates with the fidelity. Based on these findings, it is argued that there
is a need for a new sampling approach that mitigates the shift in the LIME's
framework.
| stat |
Learning Gaussian Graphical Models via Multiplicative Weights | Graphical model selection in Markov random fields is a fundamental problem in
statistics and machine learning. Two particularly prominent models, the Ising
model and Gaussian model, have largely developed in parallel using different
(though often related) techniques, and several practical algorithms with
rigorous sample complexity bounds have been established for each. In this
paper, we adapt a recently proposed algorithm of Klivans and Meka (FOCS, 2017),
based on the method of multiplicative weight updates, from the Ising model to
the Gaussian model, via non-trivial modifications to both the algorithm and its
analysis. The algorithm enjoys a sample complexity bound that is qualitatively
similar to others in the literature, has a low runtime $O(mp^2)$ in the case of
$m$ samples and $p$ nodes, and can trivially be implemented in an online
manner.
| stat |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.