A multi-scale information fusion-based multiple correlations for unsupervised attribute selection

Pengfei Zhang, Dexian Wang, Zheng Yu, Yujie Zhang, Tao Jiang, Tianrui Li

Abstract: Highlights•The two-stage processing method is used to find the optimal attribute subset.•Several multi-scale information measures are defined.•The proposed information measures is designed as an unsupervised attribute selection algorithm.•This algorithm integrates multi-scale information and considers multiple interaction issues.