Online Program Home
My Program

Abstract Details

Activity Number: 539 - SPEED: Bayesian Methods and Applications in the Life and Social Sciences
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 11:35 AM to 12:20 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #332901
Title: Bayesian Analysis of High-Dimensional Point Pattern Data Sets Using Latent Multivariate Log-Gamma Random Vectors
Author(s): Heli Gao*
Companies: Florida State University
Keywords: Bayesian ; High-Dimensional; Point Pattern; Gibbs sampler; Latent Multivariate Log-Gamma; Log-Gaussian Cox Process

We develop a new approach to analyze high-dimensional spatial point patterns. Among models for a spatial point process, the Log-Gaussian Cox Process (LGCP) is commonly used, which can be represented using a hierarchical modeling structure. However, fitting this model can be computational intensive often requires the use of Metropolis-Hastings. In particular, Bayesian inference of a continuous Cox processes often requires expensive Markov Chain Monte Carlo (MCMC) posterior simulation methods. We focus here on a new Bayesian method, and propose the use of a new class of prior distributions, which leads to conjugate full-conditional distributions within a Gibbs sampler that are computational straightforward to simulate from. We demonstrate the proposed methodology through simulated examples and an analyses based on a real dataset.

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

Back to the full JSM 2018 program