arXiv subjects

Donatello Telesca

Publications and source records attributed to Donatello Telesca.

At least 19 recordsLinked to original sources

arXiv2026

Mixed Membership Models for Multilevel Functional Data

Mixed membership models extend classical clustering by substituting the notion of uncertain membership with the notion of mixed membership. In particular, these models allow each observation to partially belong to multiple pure membership classes. We discuss mixed membership models for functional data by extending the framework to multilevel functional observations. We show how the classical multivariate Karhunen-Loeve decomposition can be translated into a simple hierarchical model for scalable and flexible expressivity of the underlying stochastic processes. The identifiability of partial membership structures is aided by the definition of a hierarchical repulsive prior on the unitary simplex. Our work is motivated and illustrated by applications to a study on functional brain imaging through electroencephalography (EEG) of children with autism spectrum disorder (ASD).

stat.ME↗

arXiv2025

Bayesian Transfer Learning for High-Dimensional Linear Regression via Adaptive Shrinkage

We introduce BLAST, Bayesian Linear regression with Adaptive Shrinkage for Transfer, a Bayesian multi-source transfer learning framework for high-dimensional linear regression. The proposed analytical framework leverages global-local shrinkage priors together with Bayesian source selection to balance information sharing and regularization. We show how Bayesian source selection allows for the extraction of the most useful data sources, while discounting biasing information that may lead to negative transfer. In this framework, both source selection and sparse regression are jointly accounted for in prediction and inference via Bayesian model averaging. The structure of our model admits efficient posterior simulation via a Metropolis-within-Gibbs sampling algorithm allowing full posterior inference for the target regression coefficients, making BLAST both computationally practical and inferentially straightforward. Our method achieves more accurate posterior inference for the target than regularization approaches based on target data alone, while offering competitive predictive performance and superior uncertainty quantification compared to current state-of-the-art transfer learning methods. We validate its effectiveness through extensive simulation studies and illustrate its analytical properties when applied to a case study on the estimation of tumor mutational burden from gene expression, using data from The Cancer Genome Atlas (TCGA).

stat.ME↗

arXiv2024

Modeling EEG Spectral Features through Warped Functional Mixed Membership Models

A common concern in the field of functional data analysis is the challenge of temporal misalignment, which is typically addressed using curve registration methods. Currently, most of these methods assume the data is governed by a single common shape or a finite mixture of population level shapes. We introduce more flexibility using mixed membership models. Individual observations are assumed to partially belong to different clusters, allowing variation across multiple functional features. We propose a Bayesian hierarchical model to estimate the underlying shapes, as well as the individual time-transformation functions and levels of membership. Motivating this work is data from EEG signals in children with autism spectrum disorder (ASD). Our method agrees with the neuroimaging literature, recovering the 1/f pink noise feature distinctly from the peak in the alpha band. Furthermore, the introduction of a regression component in the estimation of time-transformation functions quantifies the effect of age and clinical designation on the location of the peak alpha frequency (PAF).

stat.ME↗

arXiv2024

Covariate Adjusted Functional Mixed Membership Models

Mixed membership models are a flexible class of probabilistic data representations used for unsupervised and semi-supervised learning, allowing each observation to partially belong to multiple clusters or features. In this manuscript, we extend the framework of functional mixed membership models to allow for covariate-dependent adjustments. The proposed model utilizes a multivariate Karhunen-Lo\`eve decomposition, which allows for a scalable and flexible model. Within this framework, we establish a set of sufficient conditions ensuring the identifiability of the mean, covariance, and allocation structure up to a permutation of the labels. This manuscript is primarily motivated by studies on functional brain imaging through electroencephalography (EEG) of children with autism spectrum disorder (ASD). Specifically, we are interested in characterizing the heterogeneity of alpha oscillations for typically developing (TD) children and children with ASD. Since alpha oscillations are known to change as children develop, we aim to characterize the heterogeneity of alpha oscillations conditionally on the age of the child. Using the proposed framework, we were able to gain novel information on the developmental trajectories of alpha oscillations for children with ASD and how the developmental trajectories differ between TD children and children with ASD.

stat.ME↗

arXiv2024

PDXpower: A Power Analysis Tool for Experimental Design in Pre-clinical Xenograft Studies for Uncensored and Censored Outcomes

In cancer research, leveraging patient-derived xenografts (PDXs) in pre-clinical experiments is a crucial approach for assessing innovative therapeutic strategies. Addressing the inherent variability in treatment response among and within individual PDX lines is essential. However, the current literature lacks a user-friendly statistical power analysis tool capable of concurrently determining the required number of PDX lines and animals per line per treatment group in this context. In this paper, we present a simulation-based R package for sample size determination, named `\textbf{PDXpower}', which is publicly available at The Comprehensive R Archive Network \url{https://CRAN.R-project.org/package=PDXpower}. The package is designed to estimate the necessary number of both PDX lines and animals per line per treatment group for the design of a PDX experiment, whether for an uncensored outcome, or a censored time-to-event outcome. Our sample size considerations rely on two widely used analytical frameworks: the mixed effects ANOVA model for uncensored outcomes and Cox's frailty model for censored data outcomes, which effectively account for both inter-PDX variability and intra-PDX correlation in treatment response. Step-by-step illustrations for utilizing the developed package are provided, catering to scenarios with or without preliminary data.

stat.AP↗

arXiv2023

Functional Mixed Membership Models

Mixed membership models, or partial membership models, are a flexible unsupervised learning method that allows each observation to belong to multiple clusters. In this paper, we propose a Bayesian mixed membership model for functional data. By using the multivariate Karhunen-Loève theorem, we are able to derive a scalable representation of Gaussian processes that maintains data-driven learning of the covariance structure. Within this framework, we establish conditional posterior consistency given a known feature allocation matrix. Compared to previous work on mixed membership models, our proposal allows for increased modeling flexibility, with the benefit of a directly interpretable mean and covariance structure. Our work is motivated by studies in functional brain imaging through electroencephalography (EEG) of children with autism spectrum disorder (ASD). In this context, our work formalizes the clinical notion of "spectrum" in terms of feature membership proportions.

stat.ME↗

arXiv2023

Flexible Regularized Estimation in High-Dimensional Mixed Membership Models

Mixed membership models are an extension of finite mixture models, where each observation can partially belong to more than one mixture component. A probabilistic framework for mixed membership models of high-dimensional continuous data is proposed with a focus on scalability and interpretability. The novel probabilistic representation of mixed membership is based on convex combinations of dependent multivariate Gaussian random vectors. In this setting, scalability is ensured through approximations of a tensor covariance structure through multivariate eigen-approximations with adaptive regularization imposed through shrinkage priors. Conditional weak posterior consistency is established on an unconstrained model, allowing for a simple posterior sampling scheme while keeping many of the desired theoretical properties of our model. The model is motivated by two biomedical case studies: a case study on functional brain imaging of children with autism spectrum disorder (ASD) and a case study on gene expression data from breast cancer tissue. These applications highlight how the typical assumption made in cluster analysis, that each observation comes from one homogeneous subgroup, may often be restrictive in several applications, leading to unnatural interpretations of data features.

stat.ME↗

arXiv2022

Prediction & Model Evaluation for Space-Time Data

Evaluation metrics for prediction error, model selection and model averaging on space-time data are understudied and poorly understood. The absence of independent replication makes prediction ambiguous as a concept and renders evaluation procedures developed for independent data inappropriate for most space-time prediction problems. Motivated by air pollution data collected during California wildfires in 2008, this manuscript attempts a formalization of the true prediction error associated with spatial interpolation. We investigate a variety of cross-validation (CV) procedures employing both simulations and case studies to provide insight into the nature of the estimand targeted by alternative data partition strategies. Consistent with recent best practice, we find that location-based cross-validation is appropriate for estimating spatial interpolation error as in our analysis of the California wildfire data. Interestingly, commonly held notions of bias-variance trade-off of CV fold size do not trivially apply to dependent data, and we recommend leave-one-location-out (LOLO) CV as the preferred prediction error metric for spatial interpolation.

stat.ME↗

arXiv2021

Treeging

Treeging combines the flexible mean structure of regression trees with the covariance-based prediction strategy of kriging into the base learner of an ensemble prediction algorithm. In so doing, it combines the strengths of the two primary types of spatial and space-time prediction models: (1) models with flexible mean structures (often machine learning algorithms) that assume independently distributed data, and (2) kriging or Gaussian Process (GP) prediction models with rich covariance structures but simple mean structures. We investigate the predictive accuracy of treeging across a thorough and widely varied battery of spatial and space-time simulation scenarios, comparing it to ordinary kriging, random forest and ensembles of ordinary kriging base learners. Treeging performs well across the board, whereas kriging suffers when dependence is weak or in the presence of spurious covariates, and random forest suffers when the covariates are less informative. Treeging also outperforms these competitors in predicting atmospheric pollutants (ozone and PM$_{2.5}$) in several case studies. We examine sensitivity to tuning parameters (number of base learners and training data sampling proportion), finding they follow the familiar intuition of their random forest counterparts. We include a discussion of scaleability, noting that any covariance approximation techniques that expedite kriging (GP) may be similarly applied to expedite treeging.

stat.ML↗

arXiv2020

Region-Referenced Spectral Power Dynamics of EEG Signals: A Hierarchical Modeling Approach

Functional brain imaging through electroencephalography (EEG) relies upon the analysis and interpretation of high-dimensional, spatially organized time series. We propose to represent time-localized frequency domain characterizations of EEG data as region-referenced functional data. This representation is coupled with a hierarchical modeling approach to multivariate functional observations. Within this familiar setting, we discuss how several prior models relate to structural assumptions about multivariate covariance operators. An overarching modeling framework, based on infinite factorial decompositions, is finally proposed to balance flexibility and efficiency in estimation. The motivating application stems from a study of implicit auditory learning, in which typically developing (TD) children, and children with autism spectrum disorder (ASD) were exposed to a continuous speech stream. Using the proposed model, we examine differential band power dynamics as brain function is interrogated throughout the duration of a computer-controlled experiment. Our work offers a novel look at previous findings in psychiatry, and provides further insights into the understanding of ASD. Our approach to inference is fully Bayesian and implemented in a highly optimized Rcpp package.

stat.AP↗

arXiv2019

Bayesian Analysis of Multidimensional Functional Data

Multi-dimensional functional data arises in numerous modern scientific experimental and observational studies. In this paper we focus on longitudinal functional data, a structured form of multidimensional functional data. Operating within a longitudinal functional framework we aim to capture low dimensional interpretable features. We propose a computationally efficient nonparametric Bayesian method to simultaneously smooth observed data, estimate conditional functional means and functional covariance surfaces. Statistical inference is based on Monte Carlo samples from the posterior measure through adaptive blocked Gibbs sampling. Several operative characteristics associated with the proposed modeling framework are assessed comparatively in a simulated environment. We illustrate the application of our work in two case studies. The first case study involves age-specific fertility collected over time for various countries. The second case study is an implicit learning experiment in children with Autism Spectrum Disorder (ASD).

stat.ME↗

arXiv2018

Inferring Brain Signals Synchronicity from a Sample of EEG Readings

Inferring patterns of synchronous brain activity from a heterogeneous sample of electroencephalograms (EEG) is scientifically and methodologically challenging. While it is intuitively and statistically appealing to rely on readings from more than one individual in order to highlight recurrent patterns of brain activation, pooling information across subjects presents non-trivial methodological problems. We discuss some of the scientific issues associated with the understanding of synchronized neuronal activity and propose a methodological framework for statistical inference from a sample of EEG readings. Our work builds on classical contributions in time-series, clustering and functional data analysis, in an effort to reframe a challenging inferential problem in the context of familiar analytical techniques. Some attention is paid to computational issues, with a proposal based on the combination of machine learning and Bayesian techniques.

stat.AP↗

arXiv2015

Bayesian Inference for Latent Biologic Structure with Determinantal Point Processes (DPP)

We discuss the use of the determinantal point process (DPP) as a prior for latent structure in biomedical applications, where inference often centers on the interpretation of latent features as biologically or clinically meaningful structure. Typical examples include mixture models, when the terms of the mixture are meant to represent clinically meaningful subpopulations (of patients, genes, etc.). Another class of examples are feature allocation models. We propose the DPP prior as a repulsive prior on latent mixture components in the first example, and as prior on feature-specific parameters in the second case. We argue that the DPP is in general an attractive prior model for latent structure when biologically relevant interpretation of such structure is desired. We illustrate the advantages of DPP prior in three case studies, including inference in mixture models for magnetic resonance images (MRI) and for protein expression, and a feature allocation model for gene expression using data from The Cancer Genome Atlas. An important part of our argument are efficient and straightforward posterior simulation methods. We implement a variation of reversible jump Markov chain Monte Carlo simulation for inference under the DPP prior, using a density with respect to the unit rate Poisson process.

stat.ME↗

arXiv2015

A Bayesian regression tree approach to identify the effect of nanoparticles' properties on toxicity profiles

We introduce a Bayesian multiple regression tree model to characterize relationships between physico-chemical properties of nanoparticles and their in-vitro toxicity over multiple doses and times of exposure. Unlike conventional models that rely on data summaries, our model solves the low sample size issue and avoids arbitrary loss of information by combining all measurements from a general exposure experiment across doses, times of exposure, and replicates. The proposed technique integrates Bayesian trees for modeling threshold effects and interactions, and penalized B-splines for dose- and time-response surface smoothing. The resulting posterior distribution is sampled by Markov Chain Monte Carlo. This method allows for inference on a number of quantities of potential interest to substantive nanotoxicology, such as the importance of physico-chemical properties and their marginal effect on toxicity. We illustrate the application of our method to the analysis of a library of 24 nano metal oxides.

stat.AP↗

arXiv2015

Non-Local Priors for High-Dimensional Estimation

Simultaneously achieving parsimony and good predictive power in high dimensions is a main challenge in statistics. Non-local priors (NLPs) possess appealing properties for high-dimensional model choice, but their use for estimation has not been studied in detail. We show that, for regular models, Bayesian model averaging (BMA) estimates based on NLPs shrink spurious parameters either at fast polynomial or quasi-exponential rates as the sample size $n$ increases (depending on the chosen prior density). Non-spurious parameter estimates only differ from the oracle MLE by a factor of $n^{-1}$. We extend some results to linear models with dimension $p$ growing with $n$. Coupled with our theoretical investigations, we outline the constructive representation of NLPs as mixtures of truncated distributions. From a practitioners' perspective, our work enables simple posterior sampling and extending NLPs beyond previous proposals. Our results show notable high-dimensional estimation for linear models with $p>>n$ at reduced computational cost. NLPs provided lower estimation error than benchmark and hyper-g priors, SCAD and LASSO in simulations, and in gene expression data achieved higher cross-validated $R^2$ with an order of magnitude less predictors. Remarkably, these results were obtained without the need to pre-screen predictors. Our findings contribute to the debate of whether different priors should be used for estimation and model selection, showing that selection priors may actually be desirable for high-dimensional estimation.

math.ST↗

arXiv2014

Comparison of Clustering Methods for Time Course Genomic Data: Applications to Aging Effects

Time course microarray data provide insight about dynamic biological processes. While several clustering methods have been proposed for the analysis of these data structures, comparison and selection of appropriate clustering methods are seldom discussed. We compared $3$ probabilistic based clustering methods and $3$ distance based clustering methods for time course microarray data. Among probabilistic methods, we considered: smoothing spline clustering also known as model based functional data analysis (MFDA), functional clustering models for sparsely sampled data (FCM) and model-based clustering (MCLUST). Among distance based methods, we considered: weighted gene co-expression network analysis (WGCNA), clustering with dynamic time warping distance (DTW) and clustering with autocorrelation based distance (ACF). We studied these algorithms in both simulated settings and case study data. Our investigations showed that FCM performed very well when gene curves were short and sparse. DTW and WGCNA performed well when gene curves were medium or long ($>=10$ observations). SSC performed very well when there were clusters of gene curves similar to one another. Overall, ACF performed poorly in these applications. In terms of computation time, FCM, SSC and DTW were considerably slower than MCLUST and WGCNA. WGCNA outperformed MCLUST by generating more accurate and biological meaningful clustering results. WGCNA and MCLUST are the best methods among the 6 methods compared, when performance and computation time are both taken into account. WGCNA outperforms MCLUST, but MCLUST provides model based inference and uncertainty measure of clustering results.

stat.AP↗

arXiv2014

Joint Clustering and Registration of Functional Data

Curve registration and clustering are fundamental tools in the analysis of functional data. While several methods have been developed and explored for either task individually, limited work has been done to infer functional clusters and register curves simultaneously. We propose a hierarchical model for joint curve clustering and registration. Our proposal combines a Dirichlet process mixture model for clustering of common shapes, with a reproducing kernel representation of phase variability for registration. We show how inference can be carried out applying standard posterior simulation algorithms and compare our method to several alternatives in both engineered data and a benchmark analysis of the Berkeley growth data. We conclude our investigation with an application to time course gene expression.

stat.ME↗

arXiv2013

Toxicity profiling of engineered nanomaterials via multivariate dose-response surface modeling

New generation in vitro high-throughput screening (HTS) assays for the assessment of engineered nanomaterials provide an opportunity to learn how these particles interact at the cellular level, particularly in relation to injury pathways. These types of assays are often characterized by small sample sizes, high measurement error and high dimensionality, as multiple cytotoxicity outcomes are measured across an array of doses and durations of exposure. In this paper we propose a probability model for the toxicity profiling of engineered nanomaterials. A hierarchical structure is used to account for the multivariate nature of the data by modeling dependence between outcomes and thereby combining information across cytotoxicity pathways. In this framework we are able to provide a flexible surface-response model that provides inference and generalizations of various classical risk assessment parameters. We discuss applications of this model to data on eight nanoparticles evaluated in relation to four cytotoxicity parameters.

stat.AP↗