The Cuckoo Search Algorithm with Suitable Probabilistic Mutation Parameters for Global Optimization Problems

Abstract

As the complexity of optimization problems have increased over the last few decades, such as the benchmark functions established by the Congress on Evolutionary Computation-2017 (CEC-2017), the development of new optimization techniques has become evident more than previously. Modern algorithms are required because conventional algorithms are inadequate to solve complicated problems. The Nearest Neighbor Cuckoo Search (NNCS), with probabilistic mutation, is studied in this work. It is the improved cuckoo search algorithm using the topology of the nearest-neighbor population and probabilistic mutation to fix the step-size problem in a search space. The proposed algorithm can solve this problem without using any NN topology, and it provides a better result than the NNCS. The _k of 0.06 was selected for both low and high dimensional problems. The proposed method has been compared with other previously-reported algorithms such as ABC, CS, PSO, FA, GSA, GWO, MVO, MFO, QPSO, LCA, NNCS in order to investigate the improvement of efficiency over the original CS.