Learning Optimization Algorithm Combined with Psychology

Learning optimization algorithm is a swarm intelligent optimization algorithm which simulates the teaching process of teachers and the learning process of students. Due to the problems of poor stability, low optimization accuracy and easy to fall into local optimal, a new teaching and learning optimization algorithm combined with psychology theory was proposed. In the improvement of the algorithm, psychological factors of teachers and students are fully considered. In the “teaching” stage of the original teaching and learning optimization algorithm, a social psychological effect is combined, namely Pygmalion effect, which claims teachers' praise, trust and expectation for students have a kind of energy, which can change the behavior of students to make students better, and to improve the local search ability and convergence of the algorithm. In the “learning” stage, fully combined with students' psychological learning characteristics, the students are divided into A, B and C based on the average score of the class and the average score of the outstanding students who are higher than the class average. Different behavioral strategies are adopted to adjust the different types of students. And then, 10 Benchmark test functions are used to compare the optimization performance with ABC, TLBO and this algorithm, and compare the learning optimization algorithm (SLBO) combining psychology with the standard SHAP, RNN, XGBoost, K-Modes, Prophet, Levenshtein for a comparative analysis of algorithm performance, so as to verify the search ability of the algorithm, to found that the algorithm has obvious advantages in convergence accuracy, speed and robustness.