2020/04/03 by Xiaoyun Li, Chenxi Wu, Ping Li · 1 citation
Computer Science · Mathematics · #Image Retrieval and Classification Techniques #Topological and Geometric Data Analysis #Data Management and Algorithms #Feature selection #Computer science #Feature (linguistics) #Pairwise comparison #Context (archaeology) #Scheme (mathematics) #Simple (philosophy) #Data mining #Selection (genetic algorithm) #Topology (electrical circuits) #Algorithm #Theoretical computer science #Pattern recognition (psychology) #Mathematics #Artificial intelligence
paper · doi:10.1609/aaai.v34i04.5908
openalex publication_date 2020/04/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
Feature selection is an important tool to deal with high dimensional data. In unsupervised case, many popular algorithms aim at maintaining the structure of the original data. In this paper, we propose a simple and effective feature selection algorithm to enhance sample similarity preservation through a new perspective, topology preservation, which is represented by persistent diagrams from the context of computational topology. This method is designed upon a unified feature selection framework called IVFS, which is inspired by random subset method. The scheme is flexible and can handle cases where the problem is analytically intractable. The proposed algorithm is able to well preserve the pairwise distances, as well as topological patterns, of the full data. We demonstrate that our algorithm can provide satisfactory performance under a sharp sub-sampling rate, which supports efficient implementation of our proposed method to large scale datasets. Extensive experiments validate the effectiveness of the proposed feature selection scheme.