A Comparative Study of Machine Learning Techniques for Software Defect Forecasting

Abstract

The software defect prediction technique yields result that development teams may examine and further contribute to industrial results. It finds all the problematic code portions, helps software developers uncover bugs, and helps them design their testing methods with the help of the model prediction. It is essential to know what percentage of categories yield the accurate forecast for early detection. Moreover, software-defected data sets are supported and at least partially recognized due to their huge dimension. Random forests (RF) and Artificial Neural Networks (ANN) are the machine learning techniques utilized in this research. The forecast for defects is created using historical data. The outcomes showed that the artificial neural network classifier performed better than the random forest classifier.