Data & Analytics
Changepoint Detection
Also known as:
Change Point Detection
Structural Break Detection
Regime Change Detection
Updated: 2/11/2026
Detection of time points at which the statistical properties of a time series significantly change.
Quick Summary
Changepoint Detection identifies the exact moment a statistical pattern changes – ideal for campaign impact analysis.
Explanation
Methods: CUSUM, PELT, Bayesian Online Changepoint Detection (BOCPD).
Marketing Relevance
Automatically detects campaign effects, market regime changes, and product change impacts.
Common Pitfalls
False changepoints with high variance. Delay in online detection.
Origin & History
CUSUM (Page, 1954). PELT (Killick et al., 2012). Bayesian Online CPD (Adams & MacKay, 2007).
Comparisons & Differences
Changepoint Detection vs. Anomaly Detection
Changepoint finds permanent changes; Anomaly Detection finds individual outliers.