Software development projects are found to overshoot both, time and cost limits. A number of cost estimation models have been proposed  to predict development costs early in the lifecycle with the hope of managing the project well within time and budget. However, studies have reported rather high error rates of prediction even in the case of the well-known models. In this study, we use the CLONALG algorithm to improve the prediction accuracy of the Desharnais and the COCOMO 81 software development models. Our results are found to outperform the prediction accuracy produced by the feature subset selection method. Besides, the CLONALG algorithm performance is found to be faster than the conventional data mining techniques.

CLONALG, software cost models, feature subset selection.