Robust Subspace Clustering with Compressed Data

TL;DR

This paper introduces a new method called row space pursuit (RSP) for robust subspace clustering directly on compressed data, effectively handling dimension reduction while maintaining high clustering accuracy and efficiency.

Contribution

The paper proposes a novel approach, RSP, that performs accurate subspace clustering on compressed data without reconstructing original high-dimensional data.

Findings

RSP outperforms competing methods in clustering accuracy.

RSP is more computationally efficient than existing approaches.

The method successfully recovers the true row space under certain conditions.

Abstract

Dimension reduction is widely regarded as an effective way for decreasing the computation, storage and communication loads of data-driven intelligent systems, leading to a growing demand for statistical methods that allow analysis (e.g., clustering) of compressed data. We therefore study in this paper a novel problem called compressive robust subspace clustering, which is to perform robust subspace clustering with the compressed data, and which is generated by projecting the original high-dimensional data onto a lower-dimensional subspace chosen at random. Given only the compressed data and sensing matrix, the proposed method, row space pursuit (RSP), recovers the authentic row space that gives correct clustering results under certain conditions. Extensive experiments show that RSP is distinctly better than the competing methods, in terms of both clustering accuracy and computational efficiency.