MACHINE LEARNING TECHNIQUES FOR THE EVALUATION OF EFFICIENCY OF THE SOFTWARE RELIABILITY GROWTH MODELS

Omal Sahar; Muhammad Ahsan Latif; Muhammad Imran

Omal Sahar Department of Computer Science, Agriculture University, Faisalabad, Pakistan
Muhammad Ahsan Latif Department of Computer Science, Agriculture University, Faisalabad, Pakistan
Muhammad Imran Department of Industrial and Management Engineering Hanyang University Ansan, South Korea

Keywords: Software Reliability, Growth Models, Genetic Algorithm, Simulated Annealing, Multi-objective Optimization

Abstract

Efficiency is a vital factor in the domain of software. Several different approaches had been used for this purpose, but no one completely assessed the efficiency and parameters of the software reliability. In this paper, a genetic alogorithm based approach proposed, for evaluating the efficiency of the SRGMs (Software reliability growth models). Genetic algorithm (GA) is a technique in artificial intelligence for optimization and problem solving with the help of selection, crossover and mutation. The experiments were conducted on four real data sets and four different traditional models.Comparing GA based approach with other approches, i.e., simulated annealing and multiple objective optimizations using genetic algorithm, the results shows that GA based approch provides very efficient results of SRGMs as compared to other selected techniques.