Partially Bayesian Variable Selection in Classification Trees

An algorithm that dynamically incorporates expert opinion in this way has two potential advantages, each improving with the quality of the expert. First, by de-emphasizing certain subsets of variables during the estimation process, unnecessary computational activity can be avoided. Second, by giving an expert's preferred variables priority, we reduce the chance that a spurious variable will appear in the model. Hence, our resulting models are potentially more interpretable and less unstable than those generated by purely data-driven algorithms. We examine these properties in both applied and simulated contexts, and discuss potential extensions of our partially Bayesian algorithm to ensemble trees and other classification and prediction methods.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Statistics

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2015

Author dc:creator

• Noe, Douglas Alan

Contributors dc:contributor

• He, Xuming

• Statistics

Language dc:language

eng

Identifier

(MiAaPQ)AAI3242952

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/87407