Random Subspace Learning Approach to High-Dimensional Outliers Detection

TL;DR

This paper presents a new outlier detection method using random subspace learning, which is efficient for high-dimensional data and performs well compared to existing techniques.

Contribution

The paper introduces a novel random subspace learning approach for outlier detection that is computationally efficient and effective in high-dimensional, low-sample size settings.

Findings

Computationally more efficient than regularized methods in high-dimensional settings.

Performs favorably in outlier detection accuracy.

Applicable to both high-dimensional low-sample size and traditional datasets.

Abstract

We introduce and develop a novel approach to outlier detection based on adaptation of random subspace learning. Our proposed method handles both high-dimension low-sample size and traditional low-dimensional high-sample size datasets. Essentially, we avoid the computational bottleneck of techniques like minimum covariance determinant (MCD) by computing the needed determinants and associated measures in much lower dimensional subspaces. Both theoretical and computational development of our approach reveal that it is computationally more efficient than the regularized methods in high-dimensional low-sample size, and often competes favorably with existing methods as far as the percentage of correct outlier detection is concerned.