Does Comma Selection Help to Cope with Local Optima?

One hope when using non-elitism in evolutionary computation is that the ability to abandon current-best solution aids leaving local optima. To improve our understanding of this mechanism, we perform a rigorous runtime analysis basic non-elitist algorithm (EA), $$(\mu ,\lambda )$$ EA, on most benchmark function with optimum, jump function. We prove for all reasonable values parameters and problem, expected EA is, apart from lower order terms, at least as large its elitist counterpart, +\lambda (for which conduct first functions allow comparison). Consequently, leave optima inferior solutions does not lead advantage. complement bound an upper that, broad ranges parameters, identical terms. This result multi-modal problem tight