 |
Conference papers
Component elimination strategies to fit mixtures of multiple scale distributions
Abstract : We address the issue of selecting automatically the number of components in mixture models with non-Gaussian components. As a more efficient alternative to the traditional comparison of several model scores in a range, we consider procedures based on a single run of the inference scheme. Starting from an overfitting mixture in a Bayesian setting, we investigate two strategies to eliminate superfluous components. We implement these strategies for mixtures of multiple scale distributions which exhibit a variety of shapes not necessarily elliptical while remaining analytical and tractable in multiple dimensions. A Bayesian formulation and a tractable inference procedure based on variational approximation are proposed. Preliminary results on simulated and real data show promising performance in terms of model selection and computational time.
Cited literature [32 references]
https://hal.archives-ouvertes.fr/hal-02415090
Contributor : Florence Forbes <>
Submitted on : Monday, December 16, 2019 - 9:56:22 PM
Last modification on : Friday, December 4, 2020 - 9:34:18 AM
Long-term archiving on: : Tuesday, March 17, 2020 - 10:36:50 PM
Contributor : Florence Forbes <>
Submitted on : Monday, December 16, 2019 - 9:56:22 PM
Last modification on : Friday, December 4, 2020 - 9:34:18 AM
Long-term archiving on: : Tuesday, March 17, 2020 - 10:36:50 PM
File
CompElim.pdf
Files produced by the author(s)