Efficient fault-prone software modules selection based on complexity metrics
Yigang Wang, Xiaodong Zhu, Yanbin Liu Fei Ye · 2009
In order to improve the software reliability early, this paper proposes an efficient algorithm to select fault-prone software module. Based on software module's complexity metrics, the algorithm uses modified cascaded-correlation algorithm as neural network classifier to select the fault-prone software module. Finally, by analyzing the algorithm's application in the project MAP, the paper shows the advantage of the algorithm.