Neural Network Models for Agile Software Effort Estimation based on Story Points

Agile software development is now accepted as a superior alternative to conventional methods of software development, because of its inherent benefits like iterative development, rapid delivery and reduced risk. Hence, the industry must be able to efficiently estimate the effort necessary to develop projects using agile methodology. For this, different techniques like expert opinion, analogy, disaggregation etc. are adopted by researchers and practitioners. But no proper mathematical model exists for this. The existing techniques are ad-hoc and are thus prone to be incorrect. One popular approach of calculating effort of agile projects mathematically is the Story Point Approach (SPA). In this study, an effort has been made to improve the prediction accuracy of estimation done using SPA. For doing this, different types of neural networks (General Regression Neural Network (GRNN), Group Method of Data Handling (GMDH) Polynomial Neural Network and CascadeCorrelation Neural Network) are used. Finally, performance of models generated using these neural networks are compared and analyzed. Keywords— Agile Software Development, General Regression Neural Network, GMDH Polynomial Neural Network, Cascade Correlation Neural Network, Software Effort Estimation, Story Point Approach.