A Locality-sensitive hashing based instance selection method with its application to acceleration of feature selection

doi:10.1016/j.patcog.2025.112390

A Locality-sensitive hashing based instance selection method with its application to acceleration of feature selection

2025-08-30

0

PRE

AI

DOI：10.1016/j.patcog.2025.112390

Original

Original request for help

Save

Abstract

En 中文

• A novel LSH-based data partitioning method is proposed to achieve a swift division for similar instances. • The algorithm CISLSH is formulated to select the core instances from the partition of data (HashVoteBuckets). • CISLSH can be employed to largely compress the instance space used for feature selection. • CISLSH can significantly improve the computational efficiency of feature selection and ensure the effectiveness of the selected features.

A Locality-sensitive hashing based instance selection method with its application to acceleration of feature selection

Abstract

Journal

Organization