Cross-validation for change-point regression: pitfalls and solutions
This work solves a methodological problem for researchers and practitioners in statistics and data analysis by improving tuning parameter selection in change-point regression, though it is incremental as it builds on existing cross-validation and change-point methods.
The paper addresses the pitfalls of using cross-validation with squared error loss in change-point regression, which can lead to systematic errors in estimating the number and location of change-points, and proposes two solutions: using absolute error loss and modifying holdout sets, showing competitive or superior performance in numerical experiments, especially in misspecified models.
Cross-validation is the standard approach for tuning parameter selection in many non-parametric regression problems. However its use is less common in change-point regression, perhaps as its prediction error-based criterion may appear to permit small spurious changes and hence be less well-suited to estimation of the number and location of change-points. We show that in fact the problems of cross-validation with squared error loss are more severe and can lead to systematic under- or over-estimation of the number of change-points, and highly suboptimal estimation of the mean function in simple settings where changes are easily detectable. We propose two simple approaches to remedy these issues, the first involving the use of absolute error rather than squared error loss, and the second involving modifying the holdout sets used. For the latter, we provide conditions that permit consistent estimation of the number of change-points for a general change-point estimation procedure. We show these conditions are satisfied for least squares estimation using new results on its performance when supplied with the incorrect number of change-points. Numerical experiments show that our new approaches are competitive with common change-point methods using classical tuning parameter choices when error distributions are well-specified, but can substantially outperform these in misspecified models. An implementation of our methodology is available in the R package crossvalidationCP on CRAN.