Malte Londschien, Msc
- malte.londschien@ ai.ethz.ch
My research interests are in Causality and Digital Health. I aim to use ideas from causality to improve the robustness of machine learning models.
I am a Doctoral Fellow at the ETH AI Center, jointly supervised by Peter Bühlmann and Gunnar Rätsch. Prior to my PhD, I worked at QuantCo as a Data Scientist and Engineer and, during my studies, for Novartis as a Statistical Scientist. I studied Mathematics at ETH, where I received the ETH Master and Opportunity Award from the ETH Foundation, awarded based on talent and performance, and the Willi Studer Price, for graduating top of my class. Please see my personal website for further information.
Abstract We propose a novel multivariate nonparametric multiple change point detection method using classifiers. We construct a classifier log-likelihood ratio that uses class probability predictions to compare different change point configurations. We propose a computationally feasible search method that is particularly well suited for random forests, denoted by changeforest. However, the method can be paired with any classifier that yields class probability predictions, which we illustrate by also using a k-nearest neighbor classifier. We provide theoretical results motivating our choices. In a large simulation study, our proposed changeforest method achieves improved empirical performance compared to existing multivariate nonparametric change point detection methods. An efficient implementation of our method is made available for R, Python, and Rust users in the changeforest software package.
Authors Malte Londschien, Peter Bühlmann, and Solt Kovács
Submitted arXiv preprints
Change-Point Detection for Graphical Models in the Presence of Missing Values Journal of Computational and Graphical Statistics
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.
Authors Malte Londschien, Solt Kovács and Peter Bühlmann
Submitted Journal of Computational and Graphical Statistics