global optimization, differential evolution, adaptation, parameter settings, optimization competition

Success-History Based Parameter Adaptation forDifferential Evolution (SHADE) is extended. An adaptive mechanism is implemented in new SHADE4 variant where four different strategies compete to create a new trial point. The strategieshave probabilities computed proportional to their success in previous generations during the search. A strategy is selected randomlyproportional to strategies? probabilities. Novel SHADE4 variant outperforms the original SHADE algorithm in the most of CEC 2014 test problems. The results of novel SHADE4algorithm are submitted to the Special Session and Competition on Single Objective Real-Parameter Numerical Optimization (Single parameter-operator set based case) at CEC 2016.