A Streaming Feature Selection Method Based on Dynamic Feature Clustering and Particle Swarm Optimization

Xianfang Song,Hao Ma,Yong Zhang,Dunwei Gong,Yinan Guo,Ying Hu
DOI: https://doi.org/10.1109/tevc.2024.3451688
IF: 16.497
2024-01-01
IEEE Transactions on Evolutionary Computation
Abstract:Feature selection is an effective data preprocessing technique. In some practical applications, features may continuously arrive one by one or by groups, and we cannot know the exact number of features before learning. Streaming feature selection aims to remove redundant and irrelevant features from the continuously arriving features. The paper proposes a three-stage Streaming Feature Selection method based on Dynamic feature clustering and Particle Swarm Optimization (SFS-DPSO). In the first stage, an online relevance analysis is utilized to quickly remove irrelevant features, reducing the size of newly arrived feature groups. In the second stage, a dynamic feature clustering technique is employed to divide redundant features into different groups, thereby reducing the search space for subsequent evolutionary algorithms. In the third stage, a historical information-driven integer particle swarm optimization algorithm is exploited to search for optimal feature subset in the clustered feature space. The proposed algorithm is applied in 12 typical datasets with different difficulty levels and a real-word case, experimental results show that it can achieve better classification results in a reasonable time and is superior to most existing algorithms.
What problem does this paper attempt to address?