Matches in SemOpenAlex for { <https://semopenalex.org/work/W4300659681> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W4300659681 endingPage "139" @default.
- W4300659681 startingPage "129" @default.
- W4300659681 abstract "Feature importance is one of the most common explanations provided by Machine Learning (ML). However, different classification algorithms or different training sets could produce different rankings of predictive features. Thus, the quantification of differences between feature importance is crucial for assessing model trustworthiness. Rank-biased Overlap (RBO) is a similarity measure between incomplete, top-weighted and indefinite rankings, which are all characteristics of feature importance. In RBO, tuning persistence p allows to truncate rankings at any arbitrary depth, so to evaluate their overlapping size at increasing number of features. Classification of Parkinson’s disease (PD) with Explainable Boosting Machine (EBM) was chosen here as case study for introducing RBO in ML. An imbalanced dataset, 168 healthy controls (HC) and 396 PD patients, with 178 among clinical and imaging features was obtained from PPMI. Imbalanced, undersampled (K-Medoids) and oversampled (SMOTE) datasets were used for training EBMs, obtaining their respective feature importance. RBO score was calculated between ranking pairs incrementally increasing the depth by five features, from 1 to 178. All classifiers reached excellent AUC-ROC (~1) on test set, demonstrating the EBM prediction stability when trained on imbalanced datasets. RBO revealed that the maximum size of overlapping (80%) among rankings was obtained truncating at top 40 features, while their similarity decreased asymptotically to 50% when more than 45 features were considered. Thanks to RBO it was possible to demonstrate that, for the same accuracy, the more similar are the feature importance, the more stable is the model and the more reliable is the ML interpretability." @default.
- W4300659681 created "2022-10-03" @default.
- W4300659681 creator A5004029818 @default.
- W4300659681 creator A5015228993 @default.
- W4300659681 creator A5053324548 @default.
- W4300659681 date "2022-01-01" @default.
- W4300659681 modified "2023-10-11" @default.
- W4300659681 title "Introducing the Rank-Biased Overlap as Similarity Measure for Feature Importance in Explainable Machine Learning: A Case Study on Parkinson’s Disease" @default.
- W4300659681 cites W1553751323 @default.
- W4300659681 cites W1966835268 @default.
- W4300659681 cites W1971970850 @default.
- W4300659681 cites W1996796871 @default.
- W4300659681 cites W2021581601 @default.
- W4300659681 cites W2029203225 @default.
- W4300659681 cites W2033184625 @default.
- W4300659681 cites W2046945713 @default.
- W4300659681 cites W2083620785 @default.
- W4300659681 cites W2084341220 @default.
- W4300659681 cites W2140978740 @default.
- W4300659681 cites W2148143831 @default.
- W4300659681 cites W2149033360 @default.
- W4300659681 cites W2761181345 @default.
- W4300659681 cites W2948537574 @default.
- W4300659681 cites W3024442238 @default.
- W4300659681 cites W3126673261 @default.
- W4300659681 cites W3199347620 @default.
- W4300659681 cites W4220869651 @default.
- W4300659681 cites W4281551956 @default.
- W4300659681 doi "https://doi.org/10.1007/978-3-031-15037-1_11" @default.
- W4300659681 hasPublicationYear "2022" @default.
- W4300659681 type Work @default.
- W4300659681 citedByCount "4" @default.
- W4300659681 countsByYear W43006596812023 @default.
- W4300659681 crossrefType "book-chapter" @default.
- W4300659681 hasAuthorship W4300659681A5004029818 @default.
- W4300659681 hasAuthorship W4300659681A5015228993 @default.
- W4300659681 hasAuthorship W4300659681A5053324548 @default.
- W4300659681 hasConcept C103278499 @default.
- W4300659681 hasConcept C114614502 @default.
- W4300659681 hasConcept C115961682 @default.
- W4300659681 hasConcept C119857082 @default.
- W4300659681 hasConcept C124101348 @default.
- W4300659681 hasConcept C138885662 @default.
- W4300659681 hasConcept C153180895 @default.
- W4300659681 hasConcept C154945302 @default.
- W4300659681 hasConcept C164226766 @default.
- W4300659681 hasConcept C189430467 @default.
- W4300659681 hasConcept C2776401178 @default.
- W4300659681 hasConcept C2781067378 @default.
- W4300659681 hasConcept C33923547 @default.
- W4300659681 hasConcept C41008148 @default.
- W4300659681 hasConcept C41895202 @default.
- W4300659681 hasConcept C46686674 @default.
- W4300659681 hasConceptScore W4300659681C103278499 @default.
- W4300659681 hasConceptScore W4300659681C114614502 @default.
- W4300659681 hasConceptScore W4300659681C115961682 @default.
- W4300659681 hasConceptScore W4300659681C119857082 @default.
- W4300659681 hasConceptScore W4300659681C124101348 @default.
- W4300659681 hasConceptScore W4300659681C138885662 @default.
- W4300659681 hasConceptScore W4300659681C153180895 @default.
- W4300659681 hasConceptScore W4300659681C154945302 @default.
- W4300659681 hasConceptScore W4300659681C164226766 @default.
- W4300659681 hasConceptScore W4300659681C189430467 @default.
- W4300659681 hasConceptScore W4300659681C2776401178 @default.
- W4300659681 hasConceptScore W4300659681C2781067378 @default.
- W4300659681 hasConceptScore W4300659681C33923547 @default.
- W4300659681 hasConceptScore W4300659681C41008148 @default.
- W4300659681 hasConceptScore W4300659681C41895202 @default.
- W4300659681 hasConceptScore W4300659681C46686674 @default.
- W4300659681 hasLocation W43006596811 @default.
- W4300659681 hasOpenAccess W4300659681 @default.
- W4300659681 hasPrimaryLocation W43006596811 @default.
- W4300659681 hasRelatedWork W2806259446 @default.
- W4300659681 hasRelatedWork W2905433371 @default.
- W4300659681 hasRelatedWork W3160516639 @default.
- W4300659681 hasRelatedWork W4226258012 @default.
- W4300659681 hasRelatedWork W4310278675 @default.
- W4300659681 hasRelatedWork W4311431240 @default.
- W4300659681 hasRelatedWork W4312407344 @default.
- W4300659681 hasRelatedWork W4361193272 @default.
- W4300659681 hasRelatedWork W4384115502 @default.
- W4300659681 hasRelatedWork W2963326959 @default.
- W4300659681 isParatext "false" @default.
- W4300659681 isRetracted "false" @default.
- W4300659681 workType "book-chapter" @default.