Teaching-Learning-Feedback-Based Optimization

Teaching-learning-based Optimization (TLBO) is a popular meta-heuristic optimisation method that has been used in solving a number of scientific and engineering problems. In this paper, a new variant, namely Teaching-learning-feedback-based Optimization (TLFBO) is proposed. In addition to the two phases in the canonical TLBO, an additional feedback learning phase is employed to further speed up the convergence. The teacher in the previous generation is recorded and communicates with the current teacher to provide combined feedbacks to the learners and supervise the learning direction to avoid wasting computational efforts incurred in the previous generations. Numerical experiments on 10 well-known benchmark functions are conducted to evaluate the performance of the TLFBO, and experimental results show that the proposed TLFBO has a superior and competitive capability in solving continuous optimisation problems.