Robust Multivariate Control Chart for Outlier Detection Using Hierarchical Cluster Tree in SW2

Abstract

The goal of this paper is to develop a new multivariate control chart that can effectively detect potential outlier(s) in multi‐dimensional data while keeping the masking and swamping effects under control. The hierarchical clustering tree plays a central role in the proposed control chart, in an attempt to improve the Sullivan and Woodall's second method, known as the SW2 method. Historical multivariate datasets taken from the literature are used as the benchmarks to illustrate the performance of the proposed control charts in comparison to nine existing methods for outlier detection. The two criteria, the masking and swamping rates, are used as yardsticks for the evaluation purpose. An additional simulation study by means of Monte Carlo experiments further verifies that the proposed control chart that incorporates the hierarchical clustering tree performs much better in outlier detection and swamping prevention than the original SW2 and minimum volume ellipsoid methods. Copyright © 2012 John Wiley & Sons, Ltd.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.