Dataset Shift Detection in Non-stationary Environments Using EWMA Charts

Abstract

Dataset shift is a major challenge in the non-stationary environments wherein the input data distribution may change over time. Detecting the dataset shift point in the time-series data, where the distribution of time-series changes its properties, is of utmost interest. Dataset shift exists in a broad range of real-world systems. In such systems, there is a need for continuous monitoring of the process behavior and tracking the state of the shift so as to decide about initiating adaptive corrections in a timely manner. This paper presents an algorithm to detect the shift-point in a non-stationary time-series data. The proposed method detects the shift-point based on an exponentially weighted moving average (EWMA) control chart for auto-correlated observations. This algorithm is suitable to be run in real-time and monitors the data to detect the dataset shift. Its performance is evaluated through experiments using synthetic and real-world datasets. Results show that all the dataset-shifts are detected without the delay.

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 — 31 citations exist upstream.

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