10/10 – Lu Cheng, UIC
CBQB Seminar (in-person event)
October 10, 2022
12:00 PM - 1:30 PM
909 S Wolcott Ave, Chicago, IL 60612
CalendarDownload iCal File
This is an in-person event in room COMRB 8175 on west campus (directions). You can also watch seminar live here>>
Lu Cheng, PhD
Computer Science at UIC
Sequential Bias Mitigation and the Need for Causal Fairness
The increasing use of machine learning in high stakes domains such as healthcare and policing has brought the algorithmic fairness into the spotlight. The majority of research in fair machine learning has been focused on statistical-based measures that try to equalize the performance metrics (e.g., true positive rate) between different groups. Despite its simplicity, statistical fairness, which relies on correlation and passive observations, has its limitations, and it is essential to properly address causality in fairness.
In this talk, I will start with statistical-based fairness measures in the context of toxicity detection, a common task in Natural Language Processing (NLP). As social media data often arrives over time, I will introduce a fair reinforcement learning framework for this challenging sequential bias mitigation task. In the second part of this talk, I will first discuss the need for causality in fairness and then introduce a model of causal mediation analysis for deeper understanding of algorithmic fairness. I will conclude with open problems and challenges in causal fairness analysis.
Lu Cheng is an assistant professor of computer science at UIC. Her research focuses on developing algorithmic solutions for socially responsible AI using both statistical and causal methods. Lu's work has appeared in and been invited to top venues for AI (e.g., AAAI, IJCAI), data mining (e.g., KDD, WWW, WSDM), and NLP (e.g., ACL, COLING). She is the web chair of WSDM'22 and senior program committee member of AAAI'22-23. Lu was the recipient of the 2022 CS Outstanding Doctoral Student, 2021 ASU Engineering Dean's Dissertation Award, 2020 ASU Graduate Outstanding Research Award, 2021-22 ASU CIDSE Doctoral Fellowship, 2019 ASU Grace Hopper Celebration Scholarship, IBM Ph.D. Social Good Fellowship, and Visa Research Scholarship.
Sep 26, 2022
Oct 4, 2022