MEASURING RELIABILITY OF ASPECT-ORIENTED SOFTWARE USING A COMBINATION OF ARTIFICIAL NEURAL NETWORK AND IMPERIALIST COMPETITIVE ALGORITHM

Aspect-Oriented Software Engineering provides new ways to produce and deliver products and ultimately leads to reliable software. Reliability is an extremely important issue contributing to the quality of software. Thus, software engineers need proven mechanisms to determine the extent of software reliability. In this paper, a method for assessing reliability is proposed which takes advantage of a multilayer perceptron neural network. Furthermore, an imperialist competitive algorithm is used to optimize the weights to improve network performance. Finally, relying on root square mean error, the proposed approach is compared to a hybrid neural network-genetic algorithm method. The results show that the proposed approach exhibits lower error.