A Comparative Study on COCOMO II Model for Cost Estimation

Abstract

Due to its capacity to increase capital accuracy, Constructive Cost Model II (COCOMO II) is frequently chosen for predicting the cost of software projects. The accuracy level is frequently impacted by the large error value difference between COCOMO II and the real project cost. This problem can be improved by various optimization methods, such as BCO, ANN, Fuzzy, ACO, Cuckoo, and Grey Wolf optimization (GWO). Therefore, this study aimed to comparatively analyze the COCOMO II model for cost estimation. In this case, the implemented datasets were Nasa 93 and Turkish. In comparison to other optimization techniques, the results showed that COCOMO II-GWO with Fuzzy Gaussian reduced the outputs of MMRE by more than 16%. This subsequently led to the improvement of project cost-estimate accuracy levels.