Do Sophisticated Evolutionary Algorithms Perform Better than Simple Ones?

Our investigation aims to bridge the gap between theoretical and practical evolutionary algorithms. We compared the performance of a wide range of theory-driven EAs, from bare-bones algorithms like the (1+1) EA, a (2+1) GA and simple population-based algorithms to more sophisticated ones like the (1+(λ,λ)) GA and algorithms using fast (heavy-tailed) mutation operators, against sophisticated and highly effective EAs from specific applications.