Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376983072> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W4376983072 abstract "As predictive models -- e.g., from machine learning -- give likely outcomes, they may be used to reason on the effect of an intervention, a causal-inference task. The increasing complexity of health data has opened the door to a plethora of models, but also the Pandora box of model selection: which of these models yield the most valid causal estimates? Here we highlight that classic machine-learning model selection does not select the best outcome models for causal inference. Indeed, causal model selection should control both outcome errors for each individual, treated or not treated, whereas only one outcome is observed. Theoretically, simple risks used in machine learning do not control causal effects when treated and non-treated population differ too much. More elaborate risks build proxies of the causal error using ``nuisance'' re-weighting to compute it on the observed data. But does computing these nuisance adds noise to model selection? Drawing from an extensive empirical study, we outline a good causal model-selection procedure: using the so-called $Rtext{-risk}$; using flexible estimators to compute the nuisance models on the train set; and splitting out 10% of the data to compute risks." @default.
- W4376983072 created "2023-05-18" @default.
- W4376983072 creator A5045484943 @default.
- W4376983072 creator A5074733625 @default.
- W4376983072 date "2023-02-01" @default.
- W4376983072 modified "2023-09-29" @default.
- W4376983072 title "How to select predictive models for causal inference?" @default.
- W4376983072 doi "https://doi.org/10.48550/arxiv.2302.00370" @default.
- W4376983072 hasPublicationYear "2023" @default.
- W4376983072 type Work @default.
- W4376983072 citedByCount "0" @default.
- W4376983072 crossrefType "posted-content" @default.
- W4376983072 hasAuthorship W4376983072A5045484943 @default.
- W4376983072 hasAuthorship W4376983072A5074733625 @default.
- W4376983072 hasBestOaLocation W43769830721 @default.
- W4376983072 hasConcept C105795698 @default.
- W4376983072 hasConcept C11671645 @default.
- W4376983072 hasConcept C119857082 @default.
- W4376983072 hasConcept C126838900 @default.
- W4376983072 hasConcept C144237770 @default.
- W4376983072 hasConcept C148220186 @default.
- W4376983072 hasConcept C149782125 @default.
- W4376983072 hasConcept C154945302 @default.
- W4376983072 hasConcept C158600405 @default.
- W4376983072 hasConcept C177264268 @default.
- W4376983072 hasConcept C183115368 @default.
- W4376983072 hasConcept C185429906 @default.
- W4376983072 hasConcept C199360897 @default.
- W4376983072 hasConcept C2776214188 @default.
- W4376983072 hasConcept C33923547 @default.
- W4376983072 hasConcept C41008148 @default.
- W4376983072 hasConcept C71924100 @default.
- W4376983072 hasConcept C81917197 @default.
- W4376983072 hasConcept C93959086 @default.
- W4376983072 hasConceptScore W4376983072C105795698 @default.
- W4376983072 hasConceptScore W4376983072C11671645 @default.
- W4376983072 hasConceptScore W4376983072C119857082 @default.
- W4376983072 hasConceptScore W4376983072C126838900 @default.
- W4376983072 hasConceptScore W4376983072C144237770 @default.
- W4376983072 hasConceptScore W4376983072C148220186 @default.
- W4376983072 hasConceptScore W4376983072C149782125 @default.
- W4376983072 hasConceptScore W4376983072C154945302 @default.
- W4376983072 hasConceptScore W4376983072C158600405 @default.
- W4376983072 hasConceptScore W4376983072C177264268 @default.
- W4376983072 hasConceptScore W4376983072C183115368 @default.
- W4376983072 hasConceptScore W4376983072C185429906 @default.
- W4376983072 hasConceptScore W4376983072C199360897 @default.
- W4376983072 hasConceptScore W4376983072C2776214188 @default.
- W4376983072 hasConceptScore W4376983072C33923547 @default.
- W4376983072 hasConceptScore W4376983072C41008148 @default.
- W4376983072 hasConceptScore W4376983072C71924100 @default.
- W4376983072 hasConceptScore W4376983072C81917197 @default.
- W4376983072 hasConceptScore W4376983072C93959086 @default.
- W4376983072 hasLocation W43769830721 @default.
- W4376983072 hasLocation W43769830722 @default.
- W4376983072 hasOpenAccess W4376983072 @default.
- W4376983072 hasPrimaryLocation W43769830721 @default.
- W4376983072 hasRelatedWork W1593839135 @default.
- W4376983072 hasRelatedWork W2080556896 @default.
- W4376983072 hasRelatedWork W2301199057 @default.
- W4376983072 hasRelatedWork W2577514442 @default.
- W4376983072 hasRelatedWork W2949919985 @default.
- W4376983072 hasRelatedWork W2997261067 @default.
- W4376983072 hasRelatedWork W3020971956 @default.
- W4376983072 hasRelatedWork W3206864074 @default.
- W4376983072 hasRelatedWork W4288090793 @default.
- W4376983072 hasRelatedWork W3123288520 @default.
- W4376983072 isParatext "false" @default.
- W4376983072 isRetracted "false" @default.
- W4376983072 workType "article" @default.