Locality preserving projection on source code metrics for improved software maintainability
AI'06 Proceedings of the 19th Australian joint conference on Artificial Intelligence: advances in Artificial Intelligence
| Hi-index | 0.00 |
Artificial Immune Systems (AIS) are emerging machine learners, which embody the principles of natural immune systems for tackling complex real-world problems. The Artificial Immune Recognition System (AIRS) is a new kind of supervised learning AIS. Improving the quality of software products is one of the principal objectives of software engineering. It is well known that software metrics are the key tools in the software quality management. In this paper, we propose an AIRS-based method for software quality classification. We also compare our scheme with other conventional classification techniques. In addition, the Gain Ratio is employed to select relevant software metrics for classifiers. Results on the MDP benchmark dataset using the Error Rate (ER) and Average Sensitivity (AS) as the performance measures demonstrate that the AIRS is a promising method for software quality classification and the Gain Ratio-based metrics selection can considerably improve the performance of classifiers.