Modified Teaching–Learning-Based Optimization algorithm for global numerical optimization—A comparative study

Abstract

Teaching–Learning-Based Optimization (TLBO) is recently being used as a new, reliable, accurate and robust optimization technique for global optimization over continuous spaces. Few variants of TLBO have been proposed by researchers to improve the performance of the basic TLBO algorithm. In this paper the authors investigate the performance of a new variant of TLBO called modified TLBO (mTLBO) for global function optimization problems. The performance of mTLBO is compared with the state-of-the art forms of Particle Swarm Optimization (PSO), Differential Evolution (DE) and Artificial Bee Colony (ABC) algorithms. Several advanced variants of PSO, DE and ABC are considered for the comparison purpose. The suite of benchmark functions are chosen from the competition and special session on real parameter optimization under IEEE Congress on Evolutionary Computation (CEC) 2005. Statistical hypothesis testing is undertaken to demonstrate the significance of mTLBO over other investigated algorithms. Finally, the paper investigates the data clustering performance of mTLBO over other evolutionary algorithms on a few standard synthetic and artificial datasets. Results of our work reveal that mTLBO performs better than many other algorithms investigated in this work.