>Analysis of US Crime Rate Data with Generalized Linear and Hierarchical Bayesian Models

by Guo-Qiang Yang.

Abstract: Several generalized linear models and a hierarchical Bayesian model are applied to analyze the data for the crime rate for 47 states of the USA in 1960 \cite{database}. Monte Carlo Markov Chain (MCMC) is used to simulate the Bayesian model. Excellent fits are obtained. The models are expected to give better explanation for the data than the reported results and hence to assist making related decisions.

Key Words: Akaike Information Criterion (AIC), Schwarz Bayesian Criterion (SBC or BIC), generalized linear model, hierarchical Bayesian, Monte Carlo Markov Chain (MCMC)

Author:
Guo-Qiang Yang, yanggq@math.duke.edu

Editor: N. Rao Chaganty,rchagant@odu.edu

READING THE ARTICLE: You can read the article in portable document (.pdf) format (257443 bytes.)

NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.

This page has been accessed 2458 times since July 24, 2006.


Return to the InterStat Home Page.