A Hybrid Cuckoo Optimization and Harmony Search Algorithm for Software Cost Estimation

Nowadays, software effort and time estimation becomes a substantial factor in software engineering community. Accurate estimation can enhances performance in managing the project schedule, human resource allocation, cost estimation, etc. Constructive Cost Model (COCOMO) II is commonly used model of software estimation influenced by four coefficients. The optimal coeffiecients will produce optimal model in estimating effort and time development. To overcome that problem, this paper proposed a hybrid model of cuckoo search and harmony search algorithm for optimizing four coefficients of COCOMO-II to get optimal estimation. The proposed approach is applied on NASA 93 dataset and evaluated by using Magnitude of Relative Error (MRE) and Magnitude of Relative Error (MMRE). Experimental results show that the proposed method is more effective in estimating effort and time development of the software project than COCOMO-II and cuckoo search algorithm.