Adaptive and Intelligent Swarms Based Algorithm for Software Cost Estimation
Mukesh Kumar Khandelwal and Neetu Sharma
Software Cost estimation is a very important part of software development life cycle. An error in software cost (effort) estimation can cause a huge financial loss to the company. Generally, this cost of a software can be estimated by the expert who has a lot of experience in creating software, sometimes this is also done with the help of existing models that defines relationship between software size (input) and effort (output). These models are created by fixing the relationship between input and output parameters. However, there is always a need of perfect model which can improve the prediction ability of preexisting models. This model can be created by adding more suitable parameters and by tuning the performance of this model on existing value. In this paper, we have done some changes to very popular COCOMO) model and also perform tuning of parameters by a new version of PSO known as Adaptive and Intelligent Swarms based algorithm. The testing of this model is dome on 10 NASA projects. The experimental results prove that this new model has better estimation capabilities as compared to other existing cost estimations models.
Keywords: Adaptive, intelligent, PSO, stagnation, local optima, premature convergence