Using Time Series Models for Defect Prediction in Software Release Planning

James Tunnell, John Anvik
2015 Proceedings of the 27th International Conference on Software Engineering and Knowledge Engineering  
A time series model is presented that uses historical project information to predict the number of future defects, given the number of proposed features and improvements to be completed. This allows for hypothetical release plans to be compared by assessing their predicted impact on testing and defect-fixing time. We selected the VARX time series model as a reasonable approach. The accuracy of the model appeared low for a single dataset, but the error was found to be normally distributed.
doi:10.18293/seke2015-174 dblp:conf/seke/TunnellA15 fatcat:qkigbupn6bdhdfstwbg3axhc4a