Software Fault Proneness Prediction Using Genetic Based Machine Learning Techniques

摘要

This work is an attempt to propose a software replica to predict fault proneness by means of genetic based method implementing machine learning. The underlying method is collection of data from open source software, where the data will be in form of object oriented metrics. The said data would be used to create model for forecasting the faults. These techniques are known as genetic based Classifier Systems or learning classifier systems. Later in this work, there is in detail description about data collection technique and stepwise algorithm to get the results. In the end it can be concluded that these techniques can be used to make prediction model on object oriented data of software and can be useful pertaining to fault proneness prediction in the near the beginning stages in the development sequence. of any software (SDLC).