A new approach in software cost estimation using Hybrid Genetic and Artificial Bee Clony Algorithms

Nowadays, development and production of software is considered as an essential affair. To do so with a proper quality and an economic cost, a good estimation of the required time and cost is necessary; therefore, it can be said that Software Cost Estimation (SCE), which is usually a difficult affair, plays an effective role in the efficiency of an organization. The difficulty could be because of some reasons like incomprehensibility of software projects at the beginning of software production. So that, a current provided estimation procedure does not SCE and makes us apply new approaches to estimate more exactly. Therefore, in this paper, we have got a new approach based on the hybrid of both Genetic Algorithm (GA) and Artificial Bee Colony (ABC) algorithm to SCE. In this procedure a collection of different datum has been applied to form the test and training data and evaluate the function of the provided approach, MMER, MMRE, MDMRE, PRED(N), MSE, RMSE, MAE, MAPE, have been applied. Afterward the results of this procedure were compared with those acquired by models of COCOMO II, other procedures and both GA and ABC algorithm; finally, the results signified the decrease of errors in SCE.

کلیدواژه ها:

Genetic Algorithm ، Artificial Bee Colony Algorithm ، Software Cost Estimation ، Metaheuristic Algorithms