An Investigation Into a Hybrid Genetic Programming and Ant Colony Optimization Method for Credit Scoring

<p>This thesis proposes and investigates a new hybrid technique based on Genetic Programming (GP) and Ant Colony Optimization (ACO) techniques for inducing data classification rules. The proposed hybrid approach aims to improve on the accuracy of data classification rules produced by the original GP technique, which uses randomly generated initial populations. This hybrid technique relies on the ACO technique to produce the initial populations for the GP technique. To evaluate and compare their effectiveness in producing good data classification rules, GP, ACO, and hybrid techniques were implemented in the C programming language. The data classification rules were created and evaluated by executing these codes with two datasets for credit scoring problems, widely known as the Australian and German datasets, available from the Machine Learning Repository at the University of California, Irvine. The experimental results demonstrate that although all tree techniques yield similar accuracy during testing, on average, the hybrid ACO-GP approach performs better than either GP or ACO during training.</p>

Name thesis:degree_name

Computer Science - Applied Computing Track

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

TSYS School of Computer Science

Year dc:date.available

2012

Author dc:creator

• Aliehyaei, Rojin

Contributors dc:contributor

• Dr. Shamim Kahn
• Dr. Timothy G. Howard
• Dr. Lydia Ray

• Genetic Programming (GP)
• Ant Colony Optimization
• C Programming Language (ACO)
• Data Classification Rules
• Machine Learning Repository
• Computer Sciences
• Databases and Information Systems

Language dc:language

English

Repository record dc:identifier

https://csuepress.columbusstate.edu/theses_dissertations/61

OAI identifier oai:identifier

oai:csuepress.columbusstate.edu:theses_dissertations-1061