Parameter estimation of software reliability growth models using hybrid genetic algorithm

摘要: Software reliability is a quantifiable metric, which defined as the probability of software to operate without failure for particular period time in specific environment. Various SRGMs (software growth models) have been proposed predict software. The predicted by calculating attributes models. But parameters are generally nonlinear relationships, great problem finding optimal using traditional ways or techniques attribute calculation like Maximum Likelihood and least Square Estimation. following paper proposing new approach verticals SRGM hybrid genetic algorithm. This algorithm hybridization GA (Genetic Algorithm) real valued g PSO (particle swarm optimization). Each chromosome group values RGA then operators used directly modifying these chromosomes. While particle optimization different technique an alternative because its simplicity equal accuracy. has several benefits over conventional vertical calculations SRGM.