Matches in SemOpenAlex for { <https://semopenalex.org/work/W4367393752> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W4367393752 abstract "Despite the popularity of feature importance (FI) measures in interpretable machine learning, the statistical adequacy of these methods is rarely discussed. From a statistical perspective, a major distinction is between analyzing a variable's importance before and after adjusting for covariates - i.e., between $textit{marginal}$ and $textit{conditional}$ measures. Our work draws attention to this rarely acknowledged, yet crucial distinction and showcases its implications. Further, we reveal that for testing conditional FI, only few methods are available and practitioners have hitherto been severely restricted in method application due to mismatching data requirements. Most real-world data exhibits complex feature dependencies and incorporates both continuous and categorical data (mixed data). Both properties are oftentimes neglected by conditional FI measures. To fill this gap, we propose to combine the conditional predictive impact (CPI) framework with sequential knockoff sampling. The CPI enables conditional FI measurement that controls for any feature dependencies by sampling valid knockoffs - hence, generating synthetic data with similar statistical properties - for the data to be analyzed. Sequential knockoffs were deliberately designed to handle mixed data and thus allow us to extend the CPI approach to such datasets. We demonstrate through numerous simulations and a real-world example that our proposed workflow controls type I error, achieves high power and is in line with results given by other conditional FI measures, whereas marginal FI metrics result in misleading interpretations. Our findings highlight the necessity of developing statistically adequate, specialized methods for mixed data." @default.
- W4367393752 created "2023-04-30" @default.
- W4367393752 creator A5028363066 @default.
- W4367393752 creator A5067296979 @default.
- W4367393752 creator A5087084229 @default.
- W4367393752 date "2023-04-29" @default.
- W4367393752 modified "2023-10-14" @default.
- W4367393752 title "Conditional feature importance for mixed data" @default.
- W4367393752 cites W1678356000 @default.
- W4367393752 cites W2122825543 @default.
- W4367393752 cites W2132862423 @default.
- W4367393752 cites W2143891888 @default.
- W4367393752 cites W2156665896 @default.
- W4367393752 cites W2282821441 @default.
- W4367393752 cites W23724364 @default.
- W4367393752 cites W2487898712 @default.
- W4367393752 cites W2552339304 @default.
- W4367393752 cites W2625479840 @default.
- W4367393752 cites W2751886678 @default.
- W4367393752 cites W2787894218 @default.
- W4367393752 cites W2788403449 @default.
- W4367393752 cites W2893401808 @default.
- W4367393752 cites W2901837190 @default.
- W4367393752 cites W2911964244 @default.
- W4367393752 cites W2948579453 @default.
- W4367393752 cites W2963371845 @default.
- W4367393752 cites W2964060211 @default.
- W4367393752 cites W2999615587 @default.
- W4367393752 cites W3005788930 @default.
- W4367393752 cites W3033797488 @default.
- W4367393752 cites W3035517615 @default.
- W4367393752 cites W3042860357 @default.
- W4367393752 cites W3116286104 @default.
- W4367393752 cites W3158572575 @default.
- W4367393752 cites W3159263901 @default.
- W4367393752 cites W3195279903 @default.
- W4367393752 cites W3209434414 @default.
- W4367393752 cites W4226318446 @default.
- W4367393752 cites W4322716115 @default.
- W4367393752 cites W99441690 @default.
- W4367393752 doi "https://doi.org/10.1007/s10182-023-00477-9" @default.
- W4367393752 hasPublicationYear "2023" @default.
- W4367393752 type Work @default.
- W4367393752 citedByCount "0" @default.
- W4367393752 crossrefType "journal-article" @default.
- W4367393752 hasAuthorship W4367393752A5028363066 @default.
- W4367393752 hasAuthorship W4367393752A5067296979 @default.
- W4367393752 hasAuthorship W4367393752A5087084229 @default.
- W4367393752 hasBestOaLocation W43673937521 @default.
- W4367393752 hasConcept C119043178 @default.
- W4367393752 hasConcept C119857082 @default.
- W4367393752 hasConcept C124101348 @default.
- W4367393752 hasConcept C138885662 @default.
- W4367393752 hasConcept C149782125 @default.
- W4367393752 hasConcept C154945302 @default.
- W4367393752 hasConcept C177212765 @default.
- W4367393752 hasConcept C2776401178 @default.
- W4367393752 hasConcept C33923547 @default.
- W4367393752 hasConcept C41008148 @default.
- W4367393752 hasConcept C41895202 @default.
- W4367393752 hasConcept C5274069 @default.
- W4367393752 hasConcept C77088390 @default.
- W4367393752 hasConceptScore W4367393752C119043178 @default.
- W4367393752 hasConceptScore W4367393752C119857082 @default.
- W4367393752 hasConceptScore W4367393752C124101348 @default.
- W4367393752 hasConceptScore W4367393752C138885662 @default.
- W4367393752 hasConceptScore W4367393752C149782125 @default.
- W4367393752 hasConceptScore W4367393752C154945302 @default.
- W4367393752 hasConceptScore W4367393752C177212765 @default.
- W4367393752 hasConceptScore W4367393752C2776401178 @default.
- W4367393752 hasConceptScore W4367393752C33923547 @default.
- W4367393752 hasConceptScore W4367393752C41008148 @default.
- W4367393752 hasConceptScore W4367393752C41895202 @default.
- W4367393752 hasConceptScore W4367393752C5274069 @default.
- W4367393752 hasConceptScore W4367393752C77088390 @default.
- W4367393752 hasFunder F4320320879 @default.
- W4367393752 hasLocation W43673937521 @default.
- W4367393752 hasLocation W43673937522 @default.
- W4367393752 hasLocation W43673937523 @default.
- W4367393752 hasOpenAccess W4367393752 @default.
- W4367393752 hasPrimaryLocation W43673937521 @default.
- W4367393752 hasRelatedWork W1966936884 @default.
- W4367393752 hasRelatedWork W2026132343 @default.
- W4367393752 hasRelatedWork W2079796770 @default.
- W4367393752 hasRelatedWork W2171362509 @default.
- W4367393752 hasRelatedWork W2554068199 @default.
- W4367393752 hasRelatedWork W2580063968 @default.
- W4367393752 hasRelatedWork W2772283543 @default.
- W4367393752 hasRelatedWork W3124781765 @default.
- W4367393752 hasRelatedWork W4300689277 @default.
- W4367393752 hasRelatedWork W4312303673 @default.
- W4367393752 isParatext "false" @default.
- W4367393752 isRetracted "false" @default.
- W4367393752 workType "article" @default.