COCOMO II Model Parameter Optimization using PSO and Effort Estimation

Software Organizations are using a number of tools or methods for estimating the exact and accurate cost of the software project. The software industries are facing problems in its evaluation of cost of project. The challenge for software developer is to produce reliable software which is timely designed and build with proper cost estimates. Thus the methods for predicting cost of software are in the centre of attention for software industries. COCOMO II is highly a used software cost estimation model. Various techniques are employed for cost estimation such as neural logic, genetic algorithms etc. Such techniques are having difficulties like designing neural network and uncertainties in case of genetic algorithm. We propose Particle Swarm Optimization technique for optimization of COCOMO II model coefficients which can offer some crucial improvements in accuracy and can be used as a tool for effort estimation.