Software Effort Prediction - A Datamining Approach

Abstract

Effective software project estimation is one of the most challenging and important activities in software development. Proper project planning and control is not possible without a sound and reliable estimate. As a whole, the software industry does not estimate projects well and does not use estimates appropriately. We suffer far more than we should as a result and we need to focus some effort on improving the situation. Effort estimation is important to minimize the cost of a software project. The existing situation may lead to serious consequences to the company as because of poor effort estimation a major percentage of the project turns out to be either more expensive than expected, late on deliver and many more issues. Not properly giving importance to the effort estimation task by under-staffing it, running the task of low quality deliverables and setting too short schedule resulting in loss of credibility as deadlines are missed always lead to problems. The current system available for effort estimation produces non-comprehensible results. Hence the purpose of this project is to produce a software system which produces a more accurate and comprehensible results using modern tools and make it easier for the project manager to easily identify the effort needed to complete a software project in terms size of project, cost etc. The various algorithm used are Support vector machine(SVM) which are best for both classification and regression and an Active Learning Based Approach (ALBA)for rule extraction from the output of SVM to produce a comprehensible output for rule.

Keywords: Classification, Datamining, Regression.

View PDF