Finding Multivariate Outliers With FastPCS

TL;DR

This paper introduces FastPCS, an efficient algorithm for detecting multivariate outliers using the Projection Congruent Subset index, which is robust to outliers and performs well in simulations and real data applications.

Contribution

The paper presents FastPCS, a novel, fast, and affine-equivariant algorithm for multivariate outlier detection based on the new PCS index, improving robustness and efficiency.

Findings

FastPCS outperforms competitors in simulations.

FastPCS is computationally efficient.

FastPCS effectively detects outliers in real data.

Abstract

The Projection Congruent Subset (PCS) Outlyingness is a new index of multivariate outlyingness obtained by considering univariate projections of the data. Like many other outlier detection procedures, PCS searches for a subset which minimizes a criterion. The difference is that the new criterion was designed to be insensitive to the outliers. PCS is supported by FastPCS, a fast and affine equivariant algorithm which we also detail. Both an extensive simulation study and a real data application from the field of engineering show that FastPCS performs better than its competitors.