Sequential Detection of Transient Signals in High Dimensional Data Stream

TL;DR

This paper evaluates and compares the effectiveness of various sequential detection charts like EWMA, MA, CUSUM, and GLRT for identifying transient signals in high-dimensional data streams, focusing on detection power and false detection constraints.

Contribution

It provides new approximations for false detection probability and detection power, and compares the performance of these charts under different signal conditions in high-dimensional data.

Findings

EWMA performs as well as GLRT for unknown signal strength.

MEWMA with hard-threshold excels for signals in few channels.

EWMA is simple to update and does not depend on signal length.

Abstract

Motivated by sequential detection of transient signals in high dimensional data stream, we study the performance of EWMA, MA, CUSUM, and GLRT charts for detecting a transient signal in multivariate data streams in terms of the power of detection (POD) under the constraint of false detecting probability (FDP) at the stationary state. Approximations are given for FDP and POD. Comparisons show that the EWMA chart performs equally well as the GLRT chart when the signal strength is unknown, while its design is free of signal length and easy to update. In addition, the MEWMA chart with hard-threshold performs better when the signal only appears in a small portion of the channels. Dow Jones 30 industrial stock prices are used for illustration.