CS 7880: Foundations of Trustworthy Machine Learning

Seminar in Theoretical CS — Spring 2025

Date	Topic	Leader
Unit I — Privacy
Mon 01-06-25	Class 01: Intro Welcome / Introduction Privacy Overview	Jon
Thu 01-09-25	No Meeting	—
Mon 01-13-25	Class 02: Membership Inference Sankararaman, Obozinski, Jordan, Halperin. Genomic Privacy and Limits of Individual Detection in a Pool. Nature Genetics 2009 Dwork, Smith, Steinke, Ullman, Vadhan. Robust Traceability from Trace Amounts. FOCS 2015 Carlini, Chien, Nasr, Song, Terzis, Tramer. Membership Inference Attacks From First Principles. IEEE S&P 2022	Jon
Thu 01-16-25	Class 03: Differential Privacy Smith and Ullman. DPCourse Lecture 4 & 5	Jon
Mon 01-20-25	No Meeting (MLK Day)	—
Thu 01-23-25	Class 04: Differential Privacy Cont'd Smith and Ullman. DPCourse Lecture 4 & 5	Jon
Mon 01-27-25	Class 05: DP Machine Learning and Optimization Smith and Ullman. DPCourse Lecture 11 & 12 Abadi, Chu, Goodfellow, McMahan, Mironov, Talwar, Zhang. Deep Learning with Differential Privacy. CCS 2016	Hanna
Thu 01-30-25	Class 06: Memorization and Data Extraction Brown, Bun, Feldman, Smith, Talwar. When is Memorization of Irrelevant Training Data Necessary for High-Accuracy Learning. STOC 2021 Carlini et al. Extracting Training Data from Large Language Models. USENIX 2021	Tim & Mahdi
Unit II — Fairness
Mon 02-03-25	Class 07: Individual Fairness Dwork, Hardt, Pitassi, Reingold, Zemel. Fairness Through Awareness. ITCS 2012	Jack
Thu 02-06-25	Class 08: Catch Up Day	—
Mon 02-10-25	Class 09: Group and Multigroup Fairness Hardt, Price, Srebro. Equality of Opportunity in Supervised Learning. NeurIPS 2016 Kearns, Neel, Roth, Wu. Preventing Fairness Gerrymandering: Auditing and Learning for Subgroup Fairness. ICML 2018	Misha
Thu 02-13-25	Class 10: Causality and Fairness Kilbertus et al. Avoiding Discrimination Through Causal Reasoning. NeurIPS 2017	Jeff
Mon 02-17-25	No Meeting (Presidents Day)	—
Thu 02-20-25	Class 11: Causal Inference Tutorial TBD Note: Due to a scheduling conflict, we will meet in the 22nd floor conference room! I will lead a group up from the 5th floor so don't worry about access.	Jeff
Mon 02-24-25	Class 12: Fairness Tradeoffs Kleinberg, Mullainathan, Raghavan. Inherent Trade-Offs in the Fair Determination of Risk Scores. ITCS 2017	Alenna
Unit III — Calibrated Predictions
Thu 02-27-25	Class 13: Calibrated Predictions Foster and Vohra. Asymptotic Calibration. Biometrika 1998 Also see Roth's Lecture Notes	Jon
Mon 03-03-25	No Meeting (Spring Break)	—
Thu 03-06-25	No Meeting (Spring Break)	—
Mon 03-10-25	Class 14: Multicalibration Hebér-Johnson, Kim, Reingold, Rothblum. Calibration for the (Computationally-Identifiable) Masses. ICML 2018	Boce
Thu 03-13-25	Class 15: Conformal Prediction Bastani, Gupta, Jung, Noarov, Ramalingam, Roth. Practical Adversarial Multivalid Conformal Prediction. NeurIPS 2022	Zihao
Mon 03-17-25	Class 16: Omniprediction Gopalan, Kalai, Reingold, Sharan, Wieder. Omnipredictors. ITCS 2021	Ethan
Thu 03-20-25	Class 17: Calibration and Distribution Shift Kim, Kern, Goldwasser, Kreuter, Reingold. Universal adaptability: Target-independent inference that competes with propensity scoring. PNAS 2021	Willy
Mon 03-24-25	Class 18: Measuring Calibration Blasiok, Gopalan, Hu, Nakkiran. A Unifying Theory of Distance from Calibration. STOC 2023	Konstantina
Thu 03-27-25	No Meeting	—
Mon 03-31-25	Class 19: Robust Statistics and Learning Overview No reading	Jon
Thu 04-03-25	Class 20: Robust Estimation in High Dimensions Diakonikolas, Kamath, Kane, Li, Moitra, Stewart. Robust Estimators in High Dimensions without the Computational Intractability. STOC 2016 Diakonikolas, Kamath, Kane, Li, Steinhardt, Moitra. Sever: A Robust Meta-Algorithm for Stochastic Optimization. ICML 2019	Evan
Mon 04-07-25	Class 21: Certified Robustness to Data Poisoning Steinhardt, Koh, Liang. Certified Defenses for Data Poisoning Attacks. NeurIPS 2017	—
Thu 04-10-25	Class 22: Data Poisoning in Practice Carlini, Jagielski, Choquette-Choo, Paleka, Pearce, Anderson, Terzis, Thomas, Tramér. Poisoning Web-Scale Training Datasets is Practical. IEEE S&P 2024	Jenny
Mon 04-14-25	Class 23: Certified Robustness to Adversarial Examples Lecuyer, Atlidakis, Geambasu, Hsu, Jana. Certified Robustness to Adversarial Examples with Differential Privacy. IEEE S&P 2019 Optional: Cohen, Rosenfeld, Kolter. Certified Adversarial Robustness via Randomized Smoothing. ICML 2019	Alex
Thu 04-17-25	Class 24: Backdoors for Adversarial Examples Goldwasser, Kim, Vaikuntanathan, Zamir. Planting Undetectable Backdoors in Machine Learning Models FOCS 2022	—
Mon 04-21-25	No Class (Niche Sporting Event) TBD	—
Thu 04-24-25	Class 25: Adaptive Data Analysis + Baked Goods Dwork, Feldman, Hardt, Pitassi, Reingold, Roth. Preserving Statistical Validity in Adaptive Data Analysis. STOC 2015	Jon