A new semiparametric approach to finite mixture of regressions using penalized regression via fusion

Erin Austin, Wei Pan, Xiaotong Shen

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

For some modeling problems a population may be better assessed as an aggregate of unknown subpopulations, each with a distinct relationship between a response and associated variables. The finite mixture of regressions (FMR) model, in which an outcome is derived from one of a finite number of linear regression models, is a natural tool in this setting. In this article, we first propose a new penalized regression approach. Then, we demonstrate how the proposed approach better identifies subpopulations and their corresponding models than a semiparametric FMR method does. Our new method fits models for each person via grouping pursuit, utilizing a new group-truncated L1 penalty that shrinks the differences between estimated parameter vectors. The methodology causes the individuals' models to cluster into a few common models, in turn revealing previously unknown subpopulations. In fact, by varying the penalty strength, the new method can reveal a hierarchical structure among the subpopulations that can be useful in exploratory analyses. Simulations using FMR models and a real-data analysis show that the method performs promisingly well.

Original languageEnglish (US)
Pages (from-to)783-807
Number of pages25
JournalStatistica Sinica
Volume30
Issue number2
DOIs
StatePublished - Apr 2020

Keywords

  • FMR
  • Group LASSO
  • Group TLP
  • Grouping pursuit
  • Penalized regression
  • Semiparametric

Fingerprint Dive into the research topics of 'A new semiparametric approach to finite mixture of regressions using penalized regression via fusion'. Together they form a unique fingerprint.

Cite this