Bayesian Regularization based Neural Network Tool for Software Effort Estimation

Rapid growth of software industry leads to need of new technologies. Software effort estimation is one of the areas that need more concentration. Exact estimation is always a challenging task. Effort Estimation techniques are broadly classified into algorithmic and non-algorithmic techniques. An algorithmic model provides a mathematical equation for estimation which is based upon the analysis of data gathered from previously developed projects and Non-algorithmic techniques are based on new approaches, such as Soft Computing Techniques. Effective handling of cost is a basic need for any Software Organization. The main tasks for Software development estimation are determining the effort, cost and schedule of developing the project under consideration. Underestimation of project done knowingly just to win contract results into loses and also the poor quality project. So, accurate cost estimation leads to effective control of time and budget during software development. This paper presents the performance analysis of different training algorithms of neural network in effort estimation. For sake of ease, we have developed a tool in MATLAB and at last proved that Bayesian Regularization [20] gives more accurate results than other training algorithms.