FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3092876897> ?p ?o ?g. }

• W3092876897 abstract "Machine learning models in health care are often deployed in settings where it is important to protect patient privacy. In such settings, methods for differentially private (DP) learning provide a general-purpose approach to learn models with privacy guarantees. Modern methods for DP learning ensure privacy through mechanisms that censor information judged as too unique. The resulting privacy-preserving models, therefore, neglect information from the tails of a data distribution, resulting in a loss of accuracy that can disproportionately affect small groups. In this paper, we study the effects of DP learning in health care. We use state-of-the-art methods for DP learning to train privacy-preserving models in clinical prediction tasks, including x-ray classification of images and mortality prediction in time series data. We use these models to perform a comprehensive empirical investigation of the tradeoffs between privacy, utility, robustness to dataset shift, and fairness. Our results highlight lesser-known limitations of methods for DP learning in health care, models that exhibit steep tradeoffs between privacy and utility, and models whose predictions are disproportionately influenced by large demographic groups in the training data. We discuss the costs and benefits of differentially private learning in health care." @default.
• W3092876897 created "2020-10-22" @default.
• W3092876897 creator A5014472990 @default.
• W3092876897 creator A5018809423 @default.
• W3092876897 creator A5070063054 @default.
• W3092876897 creator A5087983309 @default.
• W3092876897 date "2020-10-13" @default.
• W3092876897 modified "2023-09-27" @default.
• W3092876897 title "Chasing Your Long Tails: Differentially Private Prediction in Health Care Settings" @default.
• W3092876897 cites W107524893 @default.
• W3092876897 cites W1494052777 @default.
• W3092876897 cites W1501241011 @default.
• W3092876897 cites W1580788756 @default.
• W3092876897 cites W1856966722 @default.
• W3092876897 cites W1873763122 @default.
• W3092876897 cites W1895978999 @default.
• W3092876897 cites W1986983672 @default.
• W3092876897 cites W1992129502 @default.
• W3092876897 cites W1997057722 @default.
• W3092876897 cites W2019258544 @default.
• W3092876897 cites W2027595342 @default.
• W3092876897 cites W2043254260 @default.
• W3092876897 cites W2051267297 @default.
• W3092876897 cites W2053670448 @default.
• W3092876897 cites W2054804964 @default.
• W3092876897 cites W2083878671 @default.
• W3092876897 cites W2099444409 @default.
• W3092876897 cites W2101234009 @default.
• W3092876897 cites W2104743167 @default.
• W3092876897 cites W2113189502 @default.
• W3092876897 cites W2119874464 @default.
• W3092876897 cites W2135930857 @default.
• W3092876897 cites W2136114025 @default.
• W3092876897 cites W2146544852 @default.
• W3092876897 cites W2159024459 @default.
• W3092876897 cites W2337688125 @default.
• W3092876897 cites W2396881363 @default.
• W3092876897 cites W2410097662 @default.
• W3092876897 cites W2421389337 @default.
• W3092876897 cites W2473418344 @default.
• W3092876897 cites W2520881573 @default.
• W3092876897 cites W2530395818 @default.
• W3092876897 cites W2535690855 @default.
• W3092876897 cites W2557738935 @default.
• W3092876897 cites W2597603852 @default.
• W3092876897 cites W2604834158 @default.
• W3092876897 cites W2605062226 @default.
• W3092876897 cites W2750384547 @default.
• W3092876897 cites W2751484150 @default.
• W3092876897 cites W2765693998 @default.
• W3092876897 cites W2785487418 @default.
• W3092876897 cites W2806535995 @default.
• W3092876897 cites W2886612582 @default.
• W3092876897 cites W2887355428 @default.
• W3092876897 cites W2890209530 @default.
• W3092876897 cites W2902298865 @default.
• W3092876897 cites W2902802452 @default.
• W3092876897 cites W2904188519 @default.
• W3092876897 cites W2905657479 @default.
• W3092876897 cites W2908201961 @default.
• W3092876897 cites W2914120664 @default.
• W3092876897 cites W2914514892 @default.
• W3092876897 cites W2949004837 @default.
• W3092876897 cites W2949163152 @default.
• W3092876897 cites W2950103651 @default.
• W3092876897 cites W2950220847 @default.
• W3092876897 cites W2950602864 @default.
• W3092876897 cites W2953532875 @default.
• W3092876897 cites W2962751370 @default.
• W3092876897 cites W2962968929 @default.
• W3092876897 cites W2963104135 @default.
• W3092876897 cites W2963999993 @default.
• W3092876897 cites W2964010366 @default.
• W3092876897 cites W2964256806 @default.
• W3092876897 cites W2964296660 @default.
• W3092876897 cites W2964696298 @default.
• W3092876897 cites W2967985550 @default.
• W3092876897 cites W2970631161 @default.
• W3092876897 cites W2970716886 @default.
• W3092876897 cites W2971274354 @default.
• W3092876897 cites W2971505611 @default.
• W3092876897 cites W2972199360 @default.
• W3092876897 cites W2981869278 @default.
• W3092876897 cites W2982036758 @default.
• W3092876897 cites W2982580298 @default.
• W3092876897 cites W2982749475 @default.
• W3092876897 cites W2985360425 @default.
• W3092876897 cites W2990174192 @default.
• W3092876897 cites W2992144222 @default.
• W3092876897 cites W2999251207 @default.
• W3092876897 cites W3007679811 @default.
• W3092876897 cites W3035168593 @default.
• W3092876897 cites W3035261884 @default.
• W3092876897 cites W3035493137 @default.
• W3092876897 cites W3035885149 @default.
• W3092876897 cites W3037261120 @default.
• W3092876897 cites W3037525163 @default.
• W3092876897 cites W3043023066 @default.
• W3092876897 cites W3043914326 @default.
• W3092876897 cites W3044902474 @default.

• next