FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3131911651> ?p ?o ?g. }

• W3131911651 abstract "Machine learning models are often implemented in cohort with humans in the pipeline, with the model having an option to defer to a domain expert in cases where it has low confidence in its inference. Our goal is to design mechanisms for ensuring accuracy and fairness in such prediction that combine machine learning model inferences and domain expert predictions. Prior work on systems in classification settings has focused on the setting of a pipeline with a single expert and aimed to accommodate the inaccuracies and biases of this expert to simultaneously learn an inference model and a deferral system. Our work extends this framework to settings where multiple experts are available, with each expert having their own domain of expertise and biases. We propose a framework that simultaneously learns a classifier and a deferral system, with the deferral system choosing to defer to one or more human experts in cases of input where the classifier has low confidence. We test our framework on a synthetic dataset and a content moderation dataset with biased synthetic experts, and show that it significantly improves the accuracy and fairness of the final predictions, compared to the baselines. We also collect crowdsourced labels for the content moderation task to construct a real-world dataset for the evaluation of hybrid machine-human frameworks and show that our proposed learning framework outperforms baselines on this real-world dataset as well." @default.
• W3131911651 created "2021-03-01" @default.
• W3131911651 creator A5002843568 @default.
• W3131911651 creator A5023397430 @default.
• W3131911651 creator A5080439623 @default.
• W3131911651 date "2021-02-25" @default.
• W3131911651 modified "2023-09-27" @default.
• W3131911651 title "Towards Unbiased and Accurate Deferral to Multiple Experts" @default.
• W3131911651 cites W1482468045 @default.
• W3131911651 cites W1539473175 @default.
• W3131911651 cites W1944672 @default.
• W3131911651 cites W1946594383 @default.
• W3131911651 cites W2018186689 @default.
• W3131911651 cites W2040825624 @default.
• W3131911651 cites W2070230130 @default.
• W3131911651 cites W2101946573 @default.
• W3131911651 cites W2119004707 @default.
• W3131911651 cites W2124002192 @default.
• W3131911651 cites W2141649520 @default.
• W3131911651 cites W2146502635 @default.
• W3131911651 cites W2149252982 @default.
• W3131911651 cites W2157928966 @default.
• W3131911651 cites W2174985112 @default.
• W3131911651 cites W2180771322 @default.
• W3131911651 cites W2181558882 @default.
• W3131911651 cites W2201853392 @default.
• W3131911651 cites W2250539671 @default.
• W3131911651 cites W2281090488 @default.
• W3131911651 cites W2296319761 @default.
• W3131911651 cites W2488247662 @default.
• W3131911651 cites W2522957395 @default.
• W3131911651 cites W2530395818 @default.
• W3131911651 cites W2533800767 @default.
• W3131911651 cites W2555956183 @default.
• W3131911651 cites W2586547910 @default.
• W3131911651 cites W2595653137 @default.
• W3131911651 cites W2618631885 @default.
• W3131911651 cites W2734639922 @default.
• W3131911651 cites W2757188127 @default.
• W3131911651 cites W2785327160 @default.
• W3131911651 cites W2788481061 @default.
• W3131911651 cites W2790628304 @default.
• W3131911651 cites W2891018693 @default.
• W3131911651 cites W2891947597 @default.
• W3131911651 cites W2898970033 @default.
• W3131911651 cites W2899302063 @default.
• W3131911651 cites W2901481055 @default.
• W3131911651 cites W2913410345 @default.
• W3131911651 cites W2931108066 @default.
• W3131911651 cites W2949678053 @default.
• W3131911651 cites W2962753953 @default.
• W3131911651 cites W2962755741 @default.
• W3131911651 cites W2962922829 @default.
• W3131911651 cites W2962977603 @default.
• W3131911651 cites W2963100392 @default.
• W3131911651 cites W2963178340 @default.
• W3131911651 cites W2963302453 @default.
• W3131911651 cites W2963772355 @default.
• W3131911651 cites W2963917042 @default.
• W3131911651 cites W2963917099 @default.
• W3131911651 cites W2964006987 @default.
• W3131911651 cites W2964031043 @default.
• W3131911651 cites W2970236443 @default.
• W3131911651 cites W2973205938 @default.
• W3131911651 cites W2983697263 @default.
• W3131911651 cites W2992012242 @default.
• W3131911651 cites W2996936464 @default.
• W3131911651 cites W3001754608 @default.
• W3131911651 cites W3004542466 @default.
• W3131911651 cites W3021926608 @default.
• W3131911651 cites W3021934530 @default.
• W3131911651 cites W3034220715 @default.
• W3131911651 cites W3034879135 @default.
• W3131911651 cites W3036540646 @default.
• W3131911651 cites W3037276488 @default.
• W3131911651 cites W3103146244 @default.
• W3131911651 cites W3103751997 @default.
• W3131911651 cites W3104831984 @default.
• W3131911651 cites W3181414820 @default.
• W3131911651 hasPublicationYear "2021" @default.
• W3131911651 type Work @default.
• W3131911651 sameAs 3131911651 @default.
• W3131911651 citedByCount "2" @default.
• W3131911651 countsByYear W31319116512021 @default.
• W3131911651 crossrefType "posted-content" @default.
• W3131911651 hasAuthorship W3131911651A5002843568 @default.
• W3131911651 hasAuthorship W3131911651A5023397430 @default.
• W3131911651 hasAuthorship W3131911651A5080439623 @default.
• W3131911651 hasConcept C105002631 @default.
• W3131911651 hasConcept C119857082 @default.
• W3131911651 hasConcept C121955636 @default.
• W3131911651 hasConcept C124101348 @default.
• W3131911651 hasConcept C144133560 @default.
• W3131911651 hasConcept C149782125 @default.
• W3131911651 hasConcept C154945302 @default.
• W3131911651 hasConcept C158600405 @default.
• W3131911651 hasConcept C162324750 @default.
• W3131911651 hasConcept C199360897 @default.
• W3131911651 hasConcept C207685749 @default.
• W3131911651 hasConcept C2776214188 @default.

• next