JSM 2015 Preliminary Program

Online Program Home
My Program

Abstract Details

Activity Number: 322
Type: Contributed
Date/Time: Tuesday, August 11, 2015 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistical Computing
Abstract #315901
Title: Variants of GA-Ensemble
Author(s): Dong-Yop Oh* and J. Brian Gray
Companies: The University of Texas Pan American and The University of Alabama
Keywords: AdaBoost ; classification ; GA-Ensemble ; genetic algorithm ; predictive modeling ; weak learner
Abstract:

GA-Ensemble is a method for creating a predictive ensemble model that resists outliers and reduces the model complexity by using a genetic algorithm (Oh and Gray, 2013). We propose three new variants of the original GA-Ensemble method, which initialize their starting parameters from an AdaBoost solution and then optimize using a genetic algorithm. We investigate the performance of the three GA-Ensemble variants and compare them to AdaBoost, GentleBoost, random forests, and original GA-Ensemble based on real-world data sets from the UC-Irvine Machine Learning Repository, and then the same data sets with different added noise levels. In most cases the new versions of GA-Ensemble perform better than other ensemble methods, and the first variant of GA-Ensemble is more robust to noise and provides simpler models than the others.


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

Back to the full JSM 2015 program





For program information, contact the JSM Registration Department or phone (888) 231-3473.

For Professional Development information, contact the Education Department.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

2015 JSM Online Program Home