Premium
Deconfounding and Causal Regularisation for Stability and External Validity
Author(s) -
Bühlmann Peter,
Ćevid Domagoj
Publication year - 2020
Publication title -
international statistical review
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.051
H-Index - 54
eISSN - 1751-5823
pISSN - 0306-7734
DOI - 10.1111/insr.12426
Subject(s) - robustness (evolution) , confounding , stability (learning theory) , computer science , simple (philosophy) , external validity , econometrics , causal model , causal inference , causal structure , data mining , artificial intelligence , mathematics , machine learning , statistics , epistemology , physics , quantum mechanics , biochemistry , chemistry , philosophy , gene
Summary We review some recent works on removing hidden confounding and causal regularisation 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 on the issue of concept drift, raised recently by Efron, when the data generating distribution is changing.