Change point detection for graphical models in the presence of missing values
Machine Learning
2020-10-26 v2 Machine Learning
Applications
Methodology
Abstract
We propose estimation methods for change points in high-dimensional covariance structures with an emphasis on challenging scenarios with missing values. We advocate three imputation like methods and investigate their implications on common losses used for change point detection. We also discuss how model selection methods have to be adapted to the setting of incomplete data. The methods are compared in a simulation study and applied to a time series from an environmental monitoring system. An implementation of our proposals within the R-package hdcd is available via the Supplementary materials.
Cite
@article{arxiv.1907.05409,
title = {Change point detection for graphical models in the presence of missing values},
author = {Malte Londschien and Solt Kovács and Peter Bühlmann},
journal= {arXiv preprint arXiv:1907.05409},
year = {2020}
}
Comments
14 pages, 6 figures, 3 tables, hdcd R package; added explanations and clarifications, methodology and simulation results unchanged