|
Activity Number:
|
23
- Bayesian Methods and Approaches in Big Data Analysis
|
|
Type:
|
Topic-Contributed
|
|
Date/Time:
|
Sunday, August 8, 2021 : 1:30 PM to 3:20 PM
|
|
Sponsor:
|
International Society for Bayesian Analysis (ISBA)
|
|
Abstract #317130
|
|
|
Title:
|
Avoiding Prior-Data Conflict in Regression Models via Mixture Priors
|
|
Author(s):
|
Leonardo Egidi* and Francesco Pauli and Nicola Torelli
|
|
Companies:
|
University of Trieste and University of Trieste and University of Trieste
|
|
Keywords:
|
Bayesian model;
generative model;
prior-data conflict;
regression;
regularization
|
|
Abstract:
|
The Bayesian model is constituted by the pair prior-likelihood. A prior-data conflict arises whenever the prior places most of its mass in areas of the parameter space where the likelihood is relatively low. Once a prior-data conflict is diagnosed, what to do next is a hard question to answer. We propose an automatic prior elicitation comprising of a two-component mixture of a diffuse and an informative prior distribution, weighted in such an approach to incline toward the first component if a conflict emerges. These priors result to be useful in regression models as a device for regularizing the estimates and retrieving useful inferential conclusions.
|
Authors who are presenting talks have a * after their name.
