Change Point Modelling in the Vulnerability Discovery Process

Abstract

The process of vulnerability discovery and its successful fixation is dependent on various factors like testing strategy, test case effectiveness, team constitution, efficiency, and environmental factors. These factors are prone to changes over the period of time. Change point analysis is the process of detecting this point at which the cumulative effect of factors affects the rate of change of vulnerability discovery. In this paper, we propose a mathematical model which captures point of switch or change in the regression. The practical utility of the model is confirmed by validating it on three real life software datasets. The results validate that the proposed model with change point consideration shows a better goodness of fit in comparison with mathematical models without change point.