International Journal of Computer Science and Management Studies (IJCSMS)
Effort estimation is the process of predicting the effort needed to develop software. Estimating the effort is a very crucial task in software industry. In order to get accurate estimates a lot of algorithmic models have been developed. Estimates done at the initial stage are inaccurate, where requirements are not elaborated but as the project progresses accuracy on estimate increases. So it is necessary to choose a right estimating technique. In this paper, effort estimation is calculated through a nonlinear approach Neural Network Back Propagation (NNBP). COCOMONASA dataset has been used to validate and test the effort. It is shown that neural network provides better output than most commonly used COCOMO model.