2026-09-20T03:02:59Z https://www.ideals.illinois.edu/oai-pmh

oai:www.ideals.illinois.edu:2142/124231 2024-09-16 col_2142_5131 col_2142_10761 com_2142_5130 com_2142_10755 com_2142_234

Sun, Jimeng Sun, Jimeng Rehg, James M Tong, Hanghang Romano, Yaniv 2024-05 application/pdf en text Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2024-09-16 without embargo terms The student, Zhen Lin, accepted the attached license on 2024-04-04 at 22:02. The student, Zhen Lin, submitted this Dissertation for approval on 2024-04-04 at 22:09. This Dissertation was approved for publication on 2024-04-08 at 10:35. DSpace SAF Submission Ingestion Package generated from Vireo submission #20303 on 2024-09-16 at 00:33:41 Distribution-free uncertainty quantification for deep learning Lin, Zhen 2024-04-08 Uncertainty Quantification Deep Learning Machine Learning Conformal Prediction Prediction Set Neural Networks Cross-sectional Time-series Healthcare Applications The integration of sophisticated deep learning models into critical domains, such as healthcare, autonomous vehicles, and the legal system, is increasingly becoming a trend. These models offer significant potential for enhancing outcomes efficiently, but their adoption raises crucial challenges related to model confidence, uncertainty communication, and risk management. Uncertainty Quantification (UQ) is a key framework that addresses these issues by providing a systematic way to assess and act on the reliability of model predictions. This thesis focuses on distribution-free UQ methods, which require minimal assumptions about the data distribution and model specifics, making them highly applicable across various deep learning applications. We first investigate the problem of full calibration for deep learning classifiers, aiming to address the over-confidence or under-confidence typically observed in such classifiers. Then, we discuss methods to convert a model's output into actionable uncertainty information, expressed as prediction intervals and prediction sets. In particular, we focus on the construction of prediction intervals and sets with rigorous coverage or risk control guarantees, typically provided by conformal prediction tools. Finally, we explore the problem of UQ for natural language generation (NLG), for which we apply a graph-based approach using the similarity graph of multiple sampled responses. Through exploring model calibration, risk-controlling prediction sets, conformal prediction intervals, and UQ for NLG, this thesis aims to advance the understanding and implementation of UQ in high-stakes decision-making environments. Text eng https://hdl.handle.net/2142/124231 Copyright 2024 Zhen Lin Ph.D. Dissertation Computer Science University of Illinois at Urbana-Champaign Computer Science