Select a calendar:
Filter August Events by Event Type:
Events for August 23, 2024
-
CAIS seminar: How to make optimal decisions (that are unfair, biased and non-objective)
Fri, Aug 23, 2024 @ 10:30 AM - 11:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Prof. Guido Tack, Monash University
Talk Title: How to make optimal decisions (that are unfair, biased and non-objective)
Abstract: Optimisation technology promises to help us make better decisions: plan the best route on a map, deliver goods quickly and with low emissions, construct efficient staff rosters, or design complex industrial plants. But most of these optimal decisions are in fact compromises. For example, there may be many “optimal” staff rosters that enable an organisation to function effectively and at the lowest possible cost. But some of those “optimal” rosters may be very unfair for certain staff. What if one of your staff asks you why they always get the graveyard shift, and after you’ve analysed the problem, you have to tell them it’s because their name starts with an “A”? This talk is about how optimisation technology can introduce bias and unfairness in subtle ways, and what needs to be done to fix this problem.
Biography: Guido Tack is an Associate Professor in the Department of Data Science and Artificial Intelligence at Monash University, Australia. His research focuses on combinatorial optimisation, in particular architecture and implementation techniques for constraint solvers, translation of constraint modelling languages, and industrial applications. Guido leads the development of the MiniZinc constraint modelling language and toolchain, and he is one of the leading developers of Gecode, a state-of-the-art constraint programming library. Guido’s broader research interests include programming languages and computational logic.
Host: Bistra Dilkina
More Info: https://cais.usc.edu/events/how-to-make-optimal-decisions-that-are-unfair-biased-and-non-objective/
Location: Ethel Percy Andrus Gerontology Center (GER) - 206
Audiences: Everyone Is Invited
Contact: Bistra Dilkina
Event Link: https://cais.usc.edu/events/how-to-make-optimal-decisions-that-are-unfair-biased-and-non-objective/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
-
PhD Dissertation Defense - Shushan Arakelyan
Fri, Aug 23, 2024 @ 01:00 PM - 03:00 PM
Thomas Lord Department of Computer Science
University Calendar
Title:
Building Generalizable Language Models for Code Processing
Abstract:
Successful deployment of any AI model requires generalization to previously unseen, real-world scenarios. Lack of generalization in models can lead to outcomes ranging from reduced performance to potential legal liabilities. In this thesis, I explore generalization challenges in large language models for code processing. I will discuss three different generalization concerns that language models for code processing can exhibit, and present my progress in building models that can overcome those. 1) I will start by discussing compositional generalization issues, where models must adapt to previously unseen instruction combinations 2) Next I will talk about out-of-domain generalization, and how distribution shifts within single projects or corporations can affect model performance, and how to overcome it. 3) Finally, I will talk about generalization of advanced models to programming languages with fewer resources.
Venue: SAL 213
Date/Time: August 23, 1pm-3pm
Names of the Dissertation Defense Committee members:
Xiang Ren (chair), Morteza Dehghani, Aram Galstyan, Mukund Raghothaman
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: Ellecia Williams
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
