A New Approach for Software Cost Estimation with Hybrid Genetic Algorithm and Ant Colony Optimization

• Published in: International journal of innovation and applied studies Vol. 5; no. 1; p. 72
• Main Authors: Maleki, Isa, Ghaffari, Ali, Masdari, Mohammad
• Format: Journal Article
• Language: English
• Published: Rabat International Journal of Innovation and Applied Studies 01.01.2014
• Subjects: Algorithms; Ant colony optimization; Computer programs; Genetic algorithms; Mathematical models; NASA; Software; Software development
• ISSN: 2028-9324; 2028-9324

One of the most important effective factors the software companies face is the Software Cost Estimation (SCE) in software development process time. SCE is one of the subjects which have been considered in late decades in many researches. The real estimation in software development needs effort and cost factors which are done by use of the algorithmic and Artificial Intelligence models. Boehm used the COCOMO model which is an algorithmic model in 1981 for SCE. The low accuracy and non-reliable structures of the algorithmic models led to high risks of software projects. In this article, the authors have proposed a hybrid model based on Genetic Algorithm and ACO for optimization of the effective factors' weight in NASA dataset software projects. The results of the experiments show that, the proposed model is more efficient than COCOMO model in software projects cost estimation and holds less Magnitude of Relative Error in comparison to COCOMO model.