Online Program Home
My Program

Abstract Details

Activity Number: 260 - SPEED: Topics in Bayesian Analysis
Type: Contributed
Date/Time: Monday, July 30, 2018 : 3:05 PM to 3:50 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #332892
Title: Bayesian Model Selection for Markov Chains Using Sparse Probability Vectors
Author(s): Matthew Heiner* and Athanasios Kottas and Stephan Munch
Companies: UC Santa Cruz and UC Santa Cruz and NOAA
Keywords: high order Markov chain; mixture transition distribution; nonlinear dynamics; sparsity prior

We develop a model for Bayesian selection in high order Markov chains through an extension of the mixture transition distribution of Raftery (1985). Order and lag selection are achieved through over-specification and shrinkage via novel priors for probability vectors which, in contrast to the popular Dirichlet distribution, retain sparsity properties in the presence of data. We demonstrate the model's use and utility by analyzing a time series of annual pink salmon abundance in Alaska.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program