An Improved Chicken Swarm Optimization Algorithm for Feature Selection
Haoran Wang, Zhiyu Chen, Gang Liu · Lecture notes in electrical engineering · 2022
Abstract In recent years, feature selection is becoming more and more important in data mining. Its target is that reduce the dimensionality of the datasets while at least maintaining the classification accuracy. There are some researches about chicken swarm optimization algorithm (CSO) applied to feature selection, the effect is extraordinary compared with traditional swarm intelligence algorithms. However, there is a complex search space in the challenging task feature selection, the CSO algorithm still has a default that quickly gets stuck in the local minimum problem. An improved chicken swarm optimization algorithm (ICSO) is proposed in this paper, which introduces the Levy flight strategy in the hen location update strategy and the nonlinear strategy of decreasing inertial weight in the chick location update strategy to increase the global search ability and avoid getting stuck in the local minimum problem. Compared with the other three algorithms on eighteen UCI datasets shows that the ICSO algorithm can greatly reduce the redundant features while ensuring classification accuracy.