arXiv subjects

Lorenzo Schiavon

Publications and source records attributed to Lorenzo Schiavon.

8 recordsLinked to original sources

arXiv2025

Topic-informed dynamic mixture model for occupational heterogeneity in health risk behaviors

Behavioral risk factors, i.e., smoking, poor nutrition, alcohol misuse, and physical inactivity (SNAP), are leading contributors to chronic diseases and healthcare costs worldwide. Their prevalence is shaped %not only by demographic characteristics %but and also by contextual ones such as socioeconomic and occupational environments. In this study, we leverage data from the Italian health and behavioral surveillance system PASSI to model SNAP behaviors through a Bayesian framework that integrates textual information on occupations. We use Structural Topic Modeling (STM) to cluster free-text job descriptions into latent occupational groups, which inform mixture weights in a multivariate ordered probit model. Covariate effects are allowed to vary across occupational clusters and evolve over time. To enhance interpretability and variable selection, we impose non-local spike-and-slab priors on regression coefficients. Finally, an online learning algorithm based on sequential Monte Carlo enables efficient updating as new data become available. This dynamic, scalable, and interpretable approach permits observing how occupational contexts modulate the impact of socio-demographic factors on health behaviors, providing valuable insights for targeted public health interventions.

stat.AP↗

arXiv2025

Communicating complex statistical models to a public health audience: translating science into action with the FARSI approach

Background. Effectively communicating complex statistical model outputs is a major challenge in public health. This study introduces the FARSI approach (Fast, Accessible, Reliable, Secure, Informative) as a framework to enhance the translation of intricate statistical findings into actionable insights for policymakers and stakeholders. We apply this framework in a real-world case study on chronic disease monitoring in Italy. Methods. The FARSI framework outlines key principles for developing user-friendly tools that improve the translation of statistical results. We applied these principles to create an open-access web application using R Shiny, designed to communicate chronic disease prevalence estimates from a Bayesian spatio-temporal logistic model. The case study highlights the importance of an intuitive design for fast accessibility, validated data and expert feedback for reliability, aggregated data for security, and insights into prevalence population subgroups, which were previously unobservable, for informativeness. Results. The web application enables stakeholders to explore disease prevalence across populations and geographical area through dynamic visualizations. It facilitates public health monitoring by, for instance, identifying disparities at the local level and assessing risk factors such as smoking. Its user-friendly interface enhances accessibility, making statistical findings more actionable. Conclusions. The FARSI framework provides a structured approach to improving the communication of complex research findings. By making statistical models more accessible and interpretable, it supports evidence-based decision-making in public health and increases the societal impact of research.

stat.OT↗

arXiv2024

A Bayesian approach to uncover local and temporal determinants of heterogeneity in repeated cross-sectional health surveys

In several countries, including Italy, a prominent approach to population health surveillance involves conducting repeated cross-sectional surveys at short intervals of time. These surveys gather information on the health status of individual respondents, including details on their behaviours, risk factors, and relevant socio-demographic information. While the collected data undoubtedly provides valuable information, modelling such data presents several challenges. For instance, in health risk models, it is essential to consider behavioural information, local and temporal dynamics, and disease co-occurrence. In response to these challenges, our work proposes a multivariate temporal logistic model for chronic disease diagnoses at local level. Linear predictors are modelled using individual risk factor covariates and a latent individual propensity to diseases. Leveraging a state space formulation of the model, we construct a framework in which temporal heterogeneity in regression coefficients is informed by exogenous information at local level, correspond ing to different contextual risk factors that may affect the occurrence of chronic diseases in different ways. To explore the utility and the effectiveness of our method, we analyse behavioural and risk factor surveillance data collected in Italy (PASSI), which is well-known as a country characterised by high peculiar administrative, social and territorial diversities reflected on high variability in morbidity among population subgroups.

stat.AP↗

arXiv2024

A multiscale Bayesian nonparametric framework for partial hierarchical clustering

In recent years, there has been a growing demand to discern clusters of subjects in datasets characterized by a large set of features. Often, these clusters may be highly variable in size and present partial hierarchical structures. In this context, model-based clustering approaches with nonparametric priors are gaining attention in the literature due to their flexibility and adaptability to new data. However, current approaches still face challenges in recognizing hierarchical cluster structures and in managing tiny clusters or singletons. To address these limitations, we propose a novel infinite mixture model with kernels organized within a multiscale structure. Leveraging a careful specification of the kernel parameters, our method allows the inclusion of additional information guiding possible hierarchies among clusters while maintaining flexibility. We provide theoretical support and an elegant, parsimonious formulation based on infinite factorization that allows efficient inference via Gibbs sampler.

stat.ME↗

arXiv2023

Structured factorization for single-cell gene expression data

Single-cell gene expression data are often characterized by large matrices, where the number of cells may be lower than the number of genes of interest. Factorization models have emerged as powerful tools to condense the available information through a sparse decomposition into lower rank matrices. In this work, we adapt and implement a recent Bayesian class of generalized factor models to count data and, specifically, to model the covariance between genes. The developed methodology also allows one to include exogenous information within the prior, such that recognition of covariance structures between genes is favoured. In this work, we use biological pathways as external information to induce sparsity patterns within the loadings matrix. This approach facilitates the interpretation of loadings columns and the corresponding latent factors, which can be regarded as unobserved cell covariates. We demonstrate the effectiveness of our model on single-cell RNA sequencing data obtained from lung adenocarcinoma cell lines, revealing promising insights into the role of pathways in characterizing gene relationships and extracting valuable information about unobserved cell traits.

stat.ME↗

arXiv2022

Accelerated structured matrix factorization

Matrix factorization exploits the idea that, in complex high-dimensional data, the actual signal typically lies in lower-dimensional structures. These lower dimensional objects provide useful insight, with interpretability favored by sparse structures. Sparsity, in addition, is beneficial in terms of regularization and, thus, to avoid over-fitting. By exploiting Bayesian shrinkage priors, we devise a computationally convenient approach for high-dimensional matrix factorization. The dependence between row and column entities is modeled by inducing flexible sparse patterns within factors. The availability of external information is accounted for in such a way that structures are allowed while not imposed. Inspired by boosting algorithms, we pair the the proposed approach with a numerical strategy relying on a sequential inclusion and estimation of low-rank contributions, with data-driven stopping rule. Practical advantages of the proposed approach are demonstrated by means of a simulation study and the analysis of soccer heatmaps obtained from new generation tracking data.

stat.ME↗

arXiv2022

On the dependence in football match outcomes: traditional model assumptions and an alternative proposal

The approaches routinely used to model the outcomes of football matches are characterised by strong assumptions about the dependence between the number of goals scored by the two competing teams and their marginal distribution. In this work, we argue that the assumptions traditionally made are not always based on solid arguments. Although most of these assumptions have been relaxed in the recent literature, the model introduced by Dixon and Coles in 1997 still represents a point of reference in the betting industry. While maintaining its conceptual simplicity, we propose a modification of the dependence structure. A real data application suggests that our model, named Mar-Co, outperforms the Dixon and Coles one in several betting scenarios, and parameter interpretation provides key insights on league dynamics.

stat.ME↗

arXiv2021

Generalized infinite factorization models

Factorization models express a statistical object of interest in terms of a collection of simpler objects. For example, a matrix or tensor can be expressed as a sum of rank-one components. However, in practice, it can be challenging to infer the relative impact of the different components as well as the number of components. A popular idea is to include infinitely many components having impact decreasing with the component index. This article is motivated by two limitations of existing methods: (1) lack of careful consideration of the within component sparsity structure; and (2) no accommodation for grouped variables and other non-exchangeable structures. We propose a general class of infinite factorization models that address these limitations. Theoretical support is provided, practical gains are shown in simulation studies, and an ecology application focusing on modelling bird species occurrence is discussed.

stat.ME↗