Alert button
Picture for Domagoj Ćevid

Domagoj Ćevid

Alert button

Deconfounding and Causal Regularization for Stability and External Validity

Aug 14, 2020
Peter Bühlmann, Domagoj Ćevid

Figure 1 for Deconfounding and Causal Regularization for Stability and External Validity
Figure 2 for Deconfounding and Causal Regularization for Stability and External Validity
Figure 3 for Deconfounding and Causal Regularization for Stability and External Validity
Figure 4 for Deconfounding and Causal Regularization for Stability and External Validity

We review some recent work on removing hidden confounding and causal regularization from a unified viewpoint. We describe how simple and user-friendly techniques improve stability, replicability and distributional robustness in heterogeneous data. In this sense, we provide additional thoughts to the issue on concept drift, raised by Efron (2020), when the data generating distribution is changing.

* 23 pages, 7 figures 
Viaarxiv icon

Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression

May 29, 2020
Domagoj Ćevid, Loris Michel, Nicolai Meinshausen, Peter Bühlmann

Figure 1 for Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression
Figure 2 for Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression
Figure 3 for Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression
Figure 4 for Distributional Random Forests: Heterogeneity Adjustment and Multivariate Distributional Regression

We propose an adaptation of the Random Forest algorithm to estimate the conditional distribution of a possibly multivariate response. We suggest a new splitting criterion based on the MMD two-sample test, which is suitable for detecting heterogeneity in multivariate distributions. The weights provided by the forest can be conveniently used as an input to other methods in order to locally solve various learning problems. The code is available as \texttt{R}-package \texttt{drf}.

Viaarxiv icon