Takeover time in dynamic optimization problems

Abstract

The analysis of selection pressure is a mathematical tool that has been traditionally used for studying the dynamics of population-based optimization algorithms in stationary environments, but in dynamic optimization problems (DOPs) it is still an open issue. Common metrics such as growth curve and takeover time have no clear meaning when the problem changes over time. In this article, we propose a new definition of takeover time for DOPs. For the sake of clarity, we focus on evolutionary algorithms (EA), but results could be extended to other population-based algorithms. A model for calculating takeover time values is proposed and then its accuracy is later experimentally validated.