Least squares support vector machines with genetic algorithm for estimating costs in NPD projects

Managing costs in new product development (NPD) projects is a difficult practice that requires much effort and experience. In this paper, least squares support vector machine (LS-SVM) combined with genetic algorithm (GA) are proposed to estimate cost data in these projects. The LS-SVM can overcome some shortcoming in the traditional techniques, and the GA is used to tune the LS-SVM parameters automatically. Proposed combined model can escape from the blindness of man-made choice of parameters. GA enhances the efficiency and the capability of estimation. The proposed model is more robust and reliable as compared to traditional approach. The performance of our proposed modeling approach has been tested by using the progress simulator for estimating NPD projects´ cost and compared with traditional techniques. The satisfactory results with better generalization capability and lower estimation error can be obtained.