On the Online Detection of Monotonic Trends in Time Series

Abstract

Abstract The online detection of a monotonic trend in a time series with a time‐varying mean is an important task in medical applications like intensive care monitoring, that is rendered difficult by autocorrelations. Statistical control charts designed for industrial processes are not adequate as they typically rely on a fixed target value, and many detection rules assume a trend to be linear or neglect autocorrelations. We report our experience with the online detection of slow monotonic trends. Our approach is based on a moving time window, and time‐varying autocorrelations are estimated online using parametric assumptions. The performance of versions of this approach is investigated in a simulation study. We find that shrinkage estimation of the time‐varying mean improves the results. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

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 — 23 citations exist upstream.

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