Effort estimation of software projects with optimized coefficients using soft computing technique

Software project effort measurement taking place based on elderly parameters of COCOMO II post architecture representation might not accomplish the obligatory level of exactness. Precision of results measured using the COCOMO II model will be improved by using the genetic algorithm technique. This methodology reduces the uncertainty in the COCOMO II model coefficients A, B, C and D and improves the prediction accuracy. MRE and MMRE are estimated to prove the improvement in the model accuracy.