An augmented artificial bee colony with hybrid learning

Abstract

Artificial bee colony as a recently proposed algorithm, suffers from low convergence speed when solving global optimization problems. This may due to the learning mechanism where each bee learns from the randomly selected exemplars. To address the issue, an augmented artificial bee colony algorithm, hybrid learning ABC (HLABC), is presented in this study. In HLABC, different learning strategies are adopted for the employed bee phase and the onlooker bee phase. The updating mechanism for food source position is enhanced by employing the guiding information from the global best food source. Eight benchmark functions with various properties are used to test the proposed algorithm, and the result is compared with that of original ABC, particle swarm optimization (PSO) and bacterial foraging optimization (BFO). Experimental results indicate that the designed strategy significantly improve the performance of ABC for global optimization in terms of solution accuracy and convergence speed.