Soft Computing Approach for Software Cost Estimation

Software metric and estimation is base on measuring of software attributeswhich are typically related to the product, the process and the resources ofsoftware development. One of the greatest challenges for software developersis predicting the development effort for a software system based on somemetrics for the last decades. Project managers are required to the ability togive a good estimation on software development effort. Most of the traditionaltechniques such as function points, regression models, COCOMO, etc, requirea long-term estimation process. One of the new approaches that called softcomputing techniques may offer an alternative for this challenge. This paperdescribed an enhanced soft computing model for the estimation of software costand time estimation. Result shows that the value of MMRE (Mean of Magnitudeof Relative Error) applying soft computing was substantially lower than MMREapplying by algorithmic models.