Differential Evolution Optimal Parameters Tuning with Artificial Neural Network

Differential evolution (DE) is a simple and efficient population-based stochastic algorithm for solving global numerical optimization problems. DE largely depends on parameter values search strategy. Knowledge how to tune the best of these parameters scarce. This paper aims present consistent methodology tuning optimal parameters. At heart use an artificial neural network (ANN) that learns draw links between performance values. To do so, first, data-set generated normalized, then ANN approach performed, finally, are extracted. The proposed method evaluated set 24 test problems from Black-Box Optimization Benchmarking (BBOB) benchmark. Experimental results show three distinct cases may arise with application this method. For each case, specifications about procedure follow given. Finally, comparison four rules performed in order verify validate method’s performance. study provides thorough insight into tuning, which be great users.