arXiv subjects

Bertrand Clarke

Publications and source records attributed to Bertrand Clarke.

16 recordsLinked to original sources

arXiv2026

Uncertainty Quantification Via the Posterior Predictive Variance

We use the law of total variance to generate multiple expansions for the posterior predictive variance. These expansions are sums of terms involving conditional expectations and conditional variances and provide a quantification of the sources of predictive uncertainty. Since the posterior predictive variance is fixed given the model, it represents a constant quantity that is conserved over these expansions. The terms in the expansions can be assessed in absolute or relative sense to understand the main contributors to the length of prediction intervals. We quantify the term-wise uncertainty across expansions varying in the number of terms and the order of conditionates. In particular, given that a specific term in one expansion is small or zero, we identify the other terms in other expansions that must also be small or zero. We illustrate this approach to predictive model assessment in several well-known models.

math.ST↗

arXiv2026

Changepoint Detection As Model Selection: A General Framework

This dissertation presents a general framework for changepoint detection based on L0 model selection. The core method, Iteratively Reweighted Fused Lasso (IRFL), improves upon the generalized lasso by adaptively reweighting penalties to enhance support recovery and minimize criteria such as the Bayesian Information Criterion (BIC). The approach allows for flexible modeling of seasonal patterns, linear and quadratic trends, and autoregressive dependence in the presence of changepoints. Simulation studies demonstrate that IRFL achieves accurate changepoint detection across a wide range of challenging scenarios, including those involving nuisance factors such as trends, seasonal patterns, and serially correlated errors. The framework is further extended to image data, where it enables edge-preserving denoising and segmentation, with applications spanning medical imaging and high-throughput plant phenotyping. Applications to real-world data demonstrate IRFL's utility. In particular, analysis of the Mauna Loa CO2 time series reveals changepoints that align with volcanic eruptions and ENSO events, yielding a more accurate trend decomposition than ordinary least squares. Overall, IRFL provides a robust, extensible tool for detecting structural change in complex data.

stat.ME↗

arXiv2025

Online Prediction For Streaming Observational Data

The automated collection of streaming observational data has become standard and defies most traditional analytic techniques. It is not just that models are hard to identify, there may not be any model that can be safely and usefully assumed. Indeed, frequently it is only predictions that can be made and assessed. Problems for this kind of data are often called {\cal{M}}-Open and have motivated new approaches and philosophies. This paper will review some of the most successful recent predictive methods for the {\cal{M}}-Open problem class. Techniques include predictors using expert advice such as the Shtarkov solution, Bayesian nonparametrics such as Gaussian process priors, hash function based predictors such as the {\sf Count-Min} sketch, and conformal prediction. Throughout, the properties of the predictors are presented and compared from a principled standpoint.

stat.ME↗

arXiv2024

Point Prediction for Streaming Data

We present two new approaches for point prediction with streaming data. One is based on the Count-Min sketch (CMS) and the other is based on Gaussian process priors with a random bias. These methods are intended for the most general predictive problems where no true model can be usefully formulated for the data stream. In statistical contexts, this is often called the $\mathcal{M}$-open problem class. Under the assumption that the data consists of i.i.d samples from a fixed distribution function $F$, we show that the CMS-based estimates of the distribution function are consistent. We compare our new methods with two established predictors in terms of cumulative $L^1$ error. One is based on the Shtarkov solution (often called the normalized maximum likelihood) in the normal experts setting and the other is based on Dirichlet process priors. These comparisons are for two cases. The first is one-pass meaning that the updating of the predictors is done using the fact that the CMS is a sketch. For predictors that are not one-pass, we use streaming $K$-means to give a representative subset of fixed size that can be updated as data accumulate. Preliminary computational work suggests that the one-pass median version of the CMS method is rarely outperformed by the other methods for sufficiently complex data. We also find that predictors based on Gaussian process priors with random biases perform well. The Shtarkov predictors we use here did not perform as well probably because we were only using the simplest example. The other predictors seemed to perform well mainly when the data did not look like they came from an M-open data generator.

stat.ML↗

arXiv2024

A conservation law for posterior predictive variance

We use the law of total variance to generate multiple expressions for the posterior predictive variance in Bayesian hierarchical models. These expressions are sums of terms involving conditional expectations and conditional variances. Since the posterior predictive variance is fixed given the hierarchical model, it represents a constant quantity that is conserved over the various expressions for it. The terms in the expressions can be assessed in absolute or relative terms to understand the main contributors to the length of prediction intervals. Also, sometimes these terms can be intepreted in the context of the hierarchical model. We show several examples, closed form and computational, to illustrate the uses of this approach in model assessment.

stat.ME↗

arXiv2023

Post-model-selection prediction for GLM's

We give two prediction intervals (PI) for Generalized Linear Models that take model selection uncertainty into account. The first is a straightforward extension of asymptotic normality results and the second includes an extra optimization that improves nominal coverage for small-to--moderate samples. Both PI's are wider than would be obtained without incorporating model selection uncertyainty. We compare these two PI's with three other PI's. Two are based on bootstrapping procedures and the third is based on a PI from Bayes model averaging. We argue that for general usage either the asymptotic normality or optimized asymptotic normality PI's work best. In an Appendix we extend our results to Generalized Linear Mixed Models.

stat.ME↗

arXiv2023

A Cheat Sheet for Bayesian Prediction

This paper reviews the growing field of Bayesian prediction. Bayes point and interval prediction are defined and exemplified and situated in statistical prediction more generally. Then, four general approaches to Bayes prediction are defined and we turn to predictor selection. This can be done predictively or non-predictively and predictors can be based on single models or multiple models. We call these latter cases unitary predictors and model average predictors, respectively. Then we turn to the most recent aspect of prediction to emerge, namely prediction in the context of large observational data sets and discuss three further classes of techniques. We conclude with a summary and statement of several current open problems.

stat.ME↗

arXiv2022

Testing for the Important Components of Posterior Predictive Variance

We give a decomposition of the posterior predictive variance using the law of total variance and conditioning on a finite dimensional discrete random variable. This random variable summarizes various features of modeling that are used to form the prediction for a future outcome. Then, we test which terms in this decomposition are small enough to ignore. This allows us identify which of the discrete random variables are most important to prediction intervals. The terms in the decomposition admit interpretations based on conditional means and variances and are analogous to the terms in a Cochran's theorem decomposition of squared error often used in analysis of variance. Thus, the modeling features are treated as factors in completely randomized design. In cases where there are multiple decompositions we suggest choosing the one that that gives the best predictive coverage with the smallest variance.

stat.ME↗

arXiv2022

Predictive Criteria for Prior Selection Using Shrinkage in Linear Models

Choosing a shrinkage method can be done by selecting a penalty from a list of pre-specified penalties or by constructing a penalty based on the data. If a list of penalties for a class of linear models is given, we provide comparisons based on sample size and number of non-zero parameters under a predictive stability criterion based on data perturbation. These comparisons provide recommendations for penalty selection in a variety of settings. If the preference is to construct a penalty customized for a given problem, then we propose a technique based on genetic algorithms, again using a predictive criterion. We find that, in general, a custom penalty never performs worse than any commonly used penalties but that there are cases the custom penalty reduces to a recognizable penalty. Since penalty selection is mathematically equivalent to prior selection, our method also constructs priors. The techniques and recommendations we offer are intended for finite sample cases. In this context, we argue that predictive stability under perturbation is one of the few relevant properties that can be invoked when the true model is not known. Nevertheless, we study variable inclusion in simulations and, as part of our shrinkage selection strategy, we include oracle property considerations. In particular, we see that the oracle property typically holds for penalties that satisfy basic regularity conditions and therefore is not restrictive enough to play a direct role in penalty selection. In addition, our real data example also includes considerations merging from model mis-specification.

stat.ME↗

arXiv2018

Model Selection via the VC-Dimension

We derive an objective function that can be optimized to give an estimator of the Vapnik- Chervonenkis dimension for model selection in regression problems. We verify our estimator is consistent. Then, we verify it performs well compared to seven other model selection techniques. We do this for a variety of types of data sets.

math.ST↗

arXiv2018

Modeling association in microbial communities with clique loglinear models

There is a growing awareness of the important roles that microbial communities play in complex biological processes. Modern investigation of these often uses next generation sequencing of metagenomic samples to determine community composition. We propose a statistical technique based on clique loglinear models and Bayes model averaging to identify microbial components in a metagenomic sample at various taxonomic levels that have significant associations. We describe the model class, a stochastic search technique for model selection, and the calculation of estimates of posterior probabilities of interest. We demonstrate our approach using data from the Human Microbiome Project and from a study of the skin microbiome in chronic wound healing. Our technique also identifies significant dependencies among microbial components as evidence of possible microbial syntrophy. KEYWORDS: contingency tables, graphical models, model selection, microbiome, next generation sequencing

stat.AP↗

arXiv2016

A Bayes interpretation of stacking for M-complete and M-open settings

In M-open problems where no true model can be conceptualized, it is common to back off from modeling and merely seek good prediction. Even in M-complete problems, taking a predictive approach can be very useful. Stacking is a model averaging procedure that gives a composite predictor by combining individual predictors from a list of models using weights that optimize a cross-validation criterion. We show that the stacking weights also asymptotically minimize a posterior expected loss. Hence we formally provide a Bayesian justification for cross-validation. Often the weights are constrained to be positive and sum to one. For greater generality, we omit the positivity constraint and relax the `sum to one' constraint. A key question is `What predictors should be in the average?' We first verify that the stacking error depends only on the span of the models. Then we propose using bootstrap samples from the data to generate empirical basis elements that can be used to form models. We use this in two computed examples to give stacking predictors that are (i) data driven, (ii) optimal with respect to the number of component predictors, and (iii) optimal with respect to the weight each predictor gets.

math.ST↗

arXiv2015

Clustering categorical data via ensembling dissimilarity matrices

We present a technique for clustering categorical data by generating many dissimilarity matrices and averaging over them. We begin by demonstrating our technique on low dimensional categorical data and comparing it to several other techniques that have been proposed. Then we give conditions under which our method should yield good results in general. Our method extends to high dimensional categorical data of equal lengths by ensembling over many choices of explanatory variables. In this context we compare our method with two other methods. Finally, we extend our method to high dimensional categorical data vectors of unequal length by using alignment techniques to equalize the lengths. We give examples to show that our method continues to provide good results, in particular, better in the context of genome sequences than clusterings suggested by phylogenetic trees.

stat.ML↗

arXiv2015

A General Hybrid Clustering Technique

Here, we propose a clustering technique for general clustering problems including those that have non-convex clusters. For a given desired number of clusters $K$, we use three stages to find a clustering. The first stage uses a hybrid clustering technique to produce a series of clusterings of various sizes (randomly selected). They key steps are to find a $K$-means clustering using $K_\ell$ clusters where $K_\ell \gg K$ and then joins these small clusters by using single linkage clustering. The second stage stabilizes the result of stage one by reclustering via the `membership matrix' under Hamming distance to generate a dendrogram. The third stage is to cut the dendrogram to get $K^*$ clusters where $K^* \geq K$ and then prune back to $K$ to give a final clustering. A variant on our technique also gives a reasonable estimate for $K_T$, the true number of clusters. We provide a series of arguments to justify the steps in the stages of our methods and we provide numerous examples involving real and simulated data to compare our technique with other related techniques.

stat.ML↗

arXiv2008

Pushing the Limits of Contemporary Statistics: Contributions in Honor of Jayanta K. Ghosh

Jayanta Kumar Ghosh is one of the most extraordinary professors in the field of Statistics. His research in numerous areas, especially asymptotics, has been groundbreaking, influential throughout the world, and widely recognized through awards and other honors. His leadership in Statistics as Director of the Indian Statistical Institute and President of the International Statistical Institute, among other eminent positions, has been likewise outstanding. In recognition of Jayanta's enormous impact, this volume is an effort to honor him by drawing together contributions to the main areas in which he has worked and continues to work. The papers naturally fall into five categories. First, sequential estimation was Jayanta's starting point. Thus, beginning with that topic, there are two papers, one classical by Hall and Ding leading to a variant on p-values, and one Bayesian by Berger and Sun extending reference priors to stopping time problems. Second, there are five papers in the general area of prior specification. Much of Jayanta's earlier work involved group families as does Sweeting's paper here for instance. There are also two papers dwelling on the link between fuzzy sets and priors, by Meeden and by Delampady and Angers. Equally daring is the work by Mukerjee with data dependent priors and the pleasing confluence of several prior selection criteria found by Ghosh, Santra and Kim. Jayanta himself studied a variety of prior selection criteria including probability matching priors and reference priors.

math.ST↗

arXiv2008

J. K. Ghosh's contribution to statistics: A brief outline

Professor Jayanta Kumar Ghosh has contributed massively to various areas of Statistics over the last five decades. Here, we survey some of his most important contributions. In roughly chronological order, we discuss his major results in the areas of sequential analysis, foundations, asymptotics, and Bayesian inference. It is seen that he progressed from thinking about data points, to thinking about data summarization, to the limiting cases of data summarization in as they relate to parameter estimation, and then to more general aspects of modeling including prior and model selection.

math.ST↗