Abstract

Teaching-Learning-Based Optimization (TLBO) is recently being used as a new, reliable, accurate and robust optimization technique scheme for global optimization over continuous spaces. However, the algorithm suffers from premature convergence, slow convergence rate and large computational time for optimizing the computationally expensive objective functions. Therefore, an attempt to speed up TLBO is considered necessary. This paper introduces a modification to basic TLBO that enhances the convergence rate without compromising with the solution quality. The performance of modified TLBO (mTLBO) on a test of functions is compared with original TLBO and other popular evolutionary techniques such as Artificial Bee Colony (ABC), Particle Swarm Optimization (PSO), Differential evolution (DE) etc. It is found that mTLBO requires less computational efforts to locate global optimal solution. Further, the proposed mTLBO is implemented for few well known benchmark data clustering problems and its performances compared with basic TLBO as well. Results reveal that mTLBO is able to effectively cluster data points with better cluster performance measures such as quantization errors, intra cluster and inter cluster distances compared to TLBO. This paper examines six patents on optimization using evolutionary approach and their applications to different engineering fields.

Keywords: Data clustering, function optimization, TLBO.