Institution profile

Erasmus University Rotterdam

Academic institutioneurope · nl
Official website
Research library16linked papers
Opportunities0open roles
Selected work

Representative Papers

Generalized Linear Models for Extremes: Estimation and Inference in High Dimensions

Aug 17, 2026

This study addresses the challenge of modeling extreme tail regression for high-dimensional response variables by proposing a high-dimensional extreme tail generalized linear model. By integrating Bregman divergence with L1 penalization for parameter estimation and constructing a tail-localized debiased estimator, this work establishes a unified inference framework applicable to diverse tail types. Theoretically, we prove the convergence rate and asymptotic normality of the estimator, thereby enabling valid confidence interval construction even when covariate dimensionality exceeds sample size. This approach effectively facilitates statistical inference for high-dimensional extreme data and is successfully validated through an empirical analysis of automobile insurance claims.

0 citationsRead paper

Deep Generalised Mixed Models: a Novel Neural Network Structure for Analysing Hierarchical Data

Aug 06, 2026

This study addresses the challenge of non-random missingness—such as dropout driven by negative affect—in high-dimensional experience sampling method (ESM) data, which can introduce bias in conventional approaches and standard machine learning models. The authors propose a novel neural network architecture that, for the first time, extends generalized linear mixed-effects models into a deep learning framework, jointly modeling fixed and random effects to flexibly capture both the mean structure and within-subject correlations in longitudinal data. By integrating variational autoencoders with Bayesian data augmentation, the method enables semi-parametric modeling and robust inference under general distributional assumptions and arbitrary missingness mechanisms. Empirical evaluations on the GrowIt! study and simulation experiments demonstrate its potential, though further improvements in model stability are needed to enhance practical performance.

0 citationsRead paper
Recent publications

Latest Papers

Generalized Linear Models for Extremes: Estimation and Inference in High Dimensions

Aug 17, 2026

This study addresses the challenge of modeling extreme tail regression for high-dimensional response variables by proposing a high-dimensional extreme tail generalized linear model. By integrating Bregman divergence with L1 penalization for parameter estimation and constructing a tail-localized debiased estimator, this work establishes a unified inference framework applicable to diverse tail types. Theoretically, we prove the convergence rate and asymptotic normality of the estimator, thereby enabling valid confidence interval construction even when covariate dimensionality exceeds sample size. This approach effectively facilitates statistical inference for high-dimensional extreme data and is successfully validated through an empirical analysis of automobile insurance claims.

0 citationsRead paper

Deep Generalised Mixed Models: a Novel Neural Network Structure for Analysing Hierarchical Data

Aug 06, 2026

This study addresses the challenge of non-random missingness—such as dropout driven by negative affect—in high-dimensional experience sampling method (ESM) data, which can introduce bias in conventional approaches and standard machine learning models. The authors propose a novel neural network architecture that, for the first time, extends generalized linear mixed-effects models into a deep learning framework, jointly modeling fixed and random effects to flexibly capture both the mean structure and within-subject correlations in longitudinal data. By integrating variational autoencoders with Bayesian data augmentation, the method enables semi-parametric modeling and robust inference under general distributional assumptions and arbitrary missingness mechanisms. Empirical evaluations on the GrowIt! study and simulation experiments demonstrate its potential, though further improvements in model stability are needed to enhance practical performance.

0 citationsRead paper