Applied Intelligence https://doi.org/10.1007/s10489-021-02302-9 Improved binary pigeon-inspired optimization and its application for feature selection Jeng-Shyang Pan 1 · Ai-Qing Tian 1 · Shu-Chuan Chu 1,2 · Jun-Bao Li 3 Accepted: 2 March 2021 © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2021 Abstract The Pigeon-Inspired Optimization (PIO) algorithm is an intelligent algorithm inspired by the behavior of pigeons returned to the nest. The binary pigeon-inspired optimization (BPIO) algorithm is a binary version of the PIO algorithm, it can be used to optimize binary application problems. The transfer function plays a very important part in the BPIO algorithm. To improve the solution quality of the BPIO algorithm, this paper proposes four new transfer function, an improved speed update scheme, and a second-stage position update method. The original BPIO algorithm is easier to fall into the local optimal, so a new speed update equation is proposed. In the simulation experiment, the improved BPIO is compared with binary particle swarm optimization (BPSO) and binary grey wolf optimizer (BGWO). In addition, the benchmark test function, statistical analysis, Friedman’s test and Wilcoxon rank-sum test are used to prove that the improved algorithm is quite effective, and it also verifies how to set the speed of dynamic movement. Finally, feature selection was successfully implemented in the UCI data set, and higher classification results were obtained with fewer feature numbers. Keywords Pigeon-inspired optimization · Transfer function · Binary version · Wilcoxon rank sum test · Feature selection 1 Introduction The swarm intelligence algorithm refers to a group inspired by individuals, with both the discrete behavior of individuals and the overall movement of the entire population [1, 2]. According to several simple behaviors of the individuals in the population, the entire population is affected, and  Shu-Chuan Chu scchu0803@gmail.com Jeng-Shyang Pan jspan@cc.kuas.edu.tw Ai-Qing Tian stones12138@163.com Jun-Bao Li lijunbao@hit.edu.cn 1 College of Computer Science and Engineering, Shandong University of Science and Technology, Qingdao 266590, China 2 College of Science and Engineering, Flinders University, 1284 South Road, Tonsley SA 5042, Australia 3 Harbin Institute of Technology, Harbin, China the population behaves intelligently [3, 4]. For example, the behaviors of birds, fish, and gray wolves going home, foraging, and migration, the growth behavior of microorganisms and plants. In physics, the temperature or pressure of gas molecules under the conditions of temperature and humidity and the effect of transition from stable to unstable, the movement of gas molecules always presents an optimal path. Inspired by the behaviors of biological and non-biological groups in nature, researchers have proposed a random optimization algorithm called swarm intelligence algorithm, which is used to simulate the process of individual foraging, returning home, migration, and other behaviors. It uses a random number of simulated individuals in a specified search space to simulate individuals in nature, and uses the fitness function value of the individual at a certain position as a judgment of the individual’s adaptability to the environment. According to each individual fitness can be used to simulate natural selection [5, 6]. Kenneth Sorensen proposed why the development of meta-heuristic algorithms was introduced, what are the main fallacies of most meta-heuristic algorithms, and why the field of meta-heuristic algorithms is easily affected [7]. Exhaustive search is quite impractical. No matter how clever the design, heuristic algorithms have been designed to solve