Practical Software Maintenance: Best Practices for Managing Your Software Investment
Practical Software Maintenance: Best Practices for Managing Your Software Investment
Computation of initial modes for K-modes clustering algorithm using evidence accumulation
IJCAI'07 Proceedings of the 20th international joint conference on Artifical intelligence
Clustering source code files to predict change propagation during software maintenance
Proceedings of the 50th Annual Southeast Regional Conference
| Hi-index | 0.00 |
This project addresses the issues of finding the appropriate values for the constants or parameters used in any clustering algorithm for software maintenance. It is an attempt to reduce the human effort of substituting random values in an algorithm and finding the right value for the constant by trial and error. This application implements a single objective genetic algorithm which solves the above mentioned issue in a pattern very similar to the human approach, but the computer solution is much more efficient and robust. Two clustering algorithms have also been implemented to interface with the proposed solution to study the behavior and verify the validity if the proposed solution. Experimental results show that the presented genetic-based solution is appropriate for this problem, as it tries different combinations and solutions and gives the values which in-turn helps the clustering algorithm to give optimal results.