Discrete Baby Search Algorithm for Combinatorial Optimization Problems
Yi Liu, Qibin Zheng, Gengsong Li, Jinhui Zhang, Xiaoguang Ren, Wei Qin · 2022 3rd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE) · 2022
Combinatorial optimization problems widely exist in many real-world engineering projects and artificial intelligence applications. There are many effective methods to deal with combinatorial optimization problems, of which the approaches based on evolutionary algorithms are a generally used class of methods. We proposed a novelty fabulous evolutionary algorithm called Baby Search Algorithm (BSA) which simulates the behavior of infants searching for interesting toys. BSA has a powerful ability to balance exploration and exploitation, but it is designed for handling continuous optimization issues. We develop a variant of BSA named Discrete Baby Search Algorithm (DBSA) for handling combinatorial optimization problems in this paper. There are two versions of the DBSA, one using constant value binarization strategy and the other using function discretization strategy. Besides, DBSA introduces an interesting factor to advance its performance. We take feature selection as the testing problem, and use six binary classification datasets with six classic methods to evaluate DBSA. The results demonstrate the effectiveness and superiority of our method.