• Ritu
• O. P. Sangwan

With the advent of soft-computing technologies, ecommerce and OTT platforms have gained immense market capturing capabilities. Earlier, the ability to predict the effectiveness of a campaign had to be done using expert supervision which cost a lot of money. Neural network systems have almost been successful in eliminating the need for such a specialist. This paper deals with the development of a neural network scheme for software quality prediction where a feed-forward neural network has been employed for prediction purposes. Two feed-forward neural network algorithms such as Radial Basis Function Network (RBFN) and Multilayered Perception Network (MLNN) are presented and the effectiveness of both the approaches have been analyzed. To enhance the training accuracy, a momentum term has been added in MLNN. Both the approaches are hard-coded and implemented in MATLAB where the performance of the available NN toolbox in MATLAB has also been compared with both the approaches. The comparison has been performed on 128 software dataset for which quality has to be predicted. Out of 128 dataset, 100 data pairs are used for training, and 28 data input is used for testing purposes. Furthermore, it is shown that the results obtained by the MLNN are closely matched with the actual software quality and deliver better results among all three approaches.

MATLAB, MLNN, Neural Networks, RBFN, Software Quality Prediction.

"An MLNN Based Intelligent Model for Software Quality Prediction", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 3, page no.308-318, March-2019, Available :http://www.jetir.org/papers/JETIR1903N39.pdf

"An MLNN Based Intelligent Model for Software Quality Prediction", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 3, page no. pp308-318, March-2019, Available at : http://www.jetir.org/papers/JETIR1903N39.pdf