* Faculty       * Staff       * Students & Alumni       * Committees       * Contact       * Institute Directory
* Undergraduate Program       * Graduate Program       * Courses       * Institute Catalog      
* Undergraduate       * Graduate       * Institute Admissions: Undergraduate | Graduate      
* Colloquia       * Seminars       * News       * Events       * Institute Events      
* Overview       * Lab Manual       * Institute Computing      
No Menu Selected

* News


Incentives and Inference in Multi-agent Preference Aggregation

Speaker: Lirong Xia
Center for Research on Computation and Society (CRCS), Harvard University

February 12, 2013 - 4:00 p.m. to 5:00 p.m.
Location: CII (Low) 3051
Hosted By: Dr. Elliot Anshelevich (x6491)


In many applications we need to design a system to aggregate preferences of self-interested agents. Designing such systems (called mechanisms) is usually guided by a joint consideration of agents' incentives and computational scalability, making it a challenging problem at the intersection of computer science and economics.

In this talk I will focus on social choice mechanisms that aggregate agents' preferences represented by rankings. Usually a social choice mechanism aims at achieving one of the following two goals: 1) reaching a compromise among subjective preferences, and 2) reveals the ground truth via maximum likelihood estimation (MLE) inference. I will talk about the design and analysis of social choice mechanisms on two application domains, corresponding to the two design goals respectively.

The first domain is called combinatorial voting, where the set of alternatives is exponentially large and has a combinatorial structure. I will apply game theory to analyze the effect of strategic behavior for a popular efficient mechanism called sequential voting, and show that strategic behavior of agents can be extremely harmful, which was not previously known in similar settings.

In the second domain, designing a mechanism amounts to selecting a probabilistic model, viewing agents' preferences as data, and computing the MLE. I will present an efficient EM-algorithm for a wide class of natural probabilistic models called Random Utility Models (RUMs), whose MLE inference was previously believed intractable in general. This provides a computational basis for model selection in many applications of preference aggregation in economics as well as multiagent systems, recommender systems, and crowdsourcing.


Lirong Xia is a CRCS fellow and NSF CIFellow at the Center for Research on Computation and Society (CRCS) at Harvard University. He received his Ph.D. in Computer Science in 2011 and M.A. in Economics in 2010, both from Duke University. His research focuses on the intersection of computer science and microeconomics, in particular computational social choice, game theory, mechanism design, and prediction markets. Part of his research has been reported by Duke CS news, Harvard School of Engineering and Applied Sciences News, and ACM TechNews. His Ph.D. thesis titled "Computational Voting Theory: Strategic and Combinatorial Aspects" won 2011 Duke CS outstanding Ph.D. thesis award.

Last updated: January 28, 2013