A Novel Particle Swarm Optimization Approach for Software Effort Estimation

Software Effort Estimation (SEE) is one of the main activities in development of the software projects. Effort estimation in primary stages of development of the software is one of the important challenges the software projects manager faces. One of the common models of SEE is the Constructive Cost Model (COCOMO) model. In this model, the effort for development of the software projects is a function of the Kilo Line of Code (KLOC). The use of KLOC is a very applicable factor for effort estimation in COCOMO model. Also, Artificial intelligence techniques have been applied in effort estimation very much. In this paper, we have presented a new effort estimation model for software projects using Particle Swarm Optimization (PSO) and have studied the effective parameters on effort estimation using the PSO algorithm. The results of the paper show that the proposed model gives better estimation in comparison to the COCOMO model for effort.