Generalized Linear Gaussian Cluster-Weighted Modeling

TL;DR

This paper introduces a broad family of Cluster-Weighted Models using exponential family distributions, unifying and extending existing mixture models for heterogeneous data analysis.

Contribution

It proposes Generalized Linear Gaussian Cluster Weighted Models, encompassing mixtures of generalized linear models, with theoretical and numerical validation.

Findings

Models effectively handle heterogeneous data

Mixtures of GLMs are nested within the new framework

Numerical studies demonstrate model flexibility and accuracy

Abstract

Cluster-Weighted Modeling (CWM) is a flexible mixture approach for modeling the joint probability of data coming from a heterogeneous population as a weighted sum of the products of marginal distributions and conditional distributions. In this paper, we introduce a wide family of Cluster Weighted models in which the conditional distributions are assumed to belong to the exponential family with canonical links which will be referred to as Generalized Linear Gaussian Cluster Weighted Models. Moreover, we show that, in a suitable sense, mixtures of generalized linear models can be considered as nested in Generalized Linear Gaussian Cluster Weighted Models. The proposal is illustrated through many numerical studies based on both simulated and real data sets.