Training genetic programming classifiers by vicinal-risk minimization

Ni, Ji and Rockett, Peter (2015) Training genetic programming classifiers by vicinal-risk minimization. Genetic Programming and Evolvable Machine, 16 (1). pp. 3-25. ISSN 1389-2576

Documents
GPVRM.pdf

Request a copy
[img] PDF
GPVRM.pdf - Whole Document
Restricted to Repository staff only

899kB
Item Type:Article
Item Status:Live Archive

Abstract

We propose and motivate the use of vicinal-risk minimization (VRM) for training genetic programming classifiers. We demonstrate that VRM has a number of attractive properties and demonstrate that it has a better correlation with
generalization error compared to empirical risk minimization (ERM) so is more likely to lead to better generalization performance, in general. From the results of statistical tests over a range of real and synthetic datasets, we further demonstrate that VRM yields consistently superior generalization errors compared to conventional ERM.

Keywords:Genetic programming, Classification, Vicinal-risk minimization, JCNotOpen
Subjects:G Mathematical and Computer Sciences > G700 Artificial Intelligence
G Mathematical and Computer Sciences > G760 Machine Learning
G Mathematical and Computer Sciences > G340 Statistical Modelling
Divisions:College of Science > School of Computer Science
Related URLs:
ID Code:19206
Deposited On:23 Oct 2015 14:34

Repository Staff Only: item control page