notice
March 24, 2015: Invited Speaker Seminar: From Big Data to No Data
Dr. Jia Yuan YU
Tuesday, March 24, 2015 at 10:00 a.m.
Room EV002.184
Abstract
The goal of the INSIGHT European project (insight-ict.eu) is to reduce road congestion and improve emergency response by building a decision-support system for control-room operators. This system, deployed at Dublin city council and a German federal agency, analyses large amounts of real-time data from heterogeneous sources (dedicated sensors and social media), crowdsourcing tasks to smartphone users, and outputs warnings. As an additional outcome of the project, we discover that even in the era of big data, it is possible to solve important problems with little or no communication. For instance, we develop a scheme for allocating a divisible resource (e.g., a common good such as clean air, road-access) among an unknown number of users, who do not reveal anything about their utility functions. This scheme is simple: each user iteratively adjusts its demand based on an additive-increase multiplicative-decrease algorithm. If every user follows this algorithm, then the resulting demand-profile not only converges to a social optimum, but does so while preserving user-privacy.
Biography
Jia Yuan Yu is a scientist in the Smarter Cities Lab at IBM Research Ireland, working on problems at the intersection of smart cities, data science, and multi-agent systems. One of his ongoing projects aims at reducing road congestion by deploying a decision-support system analyzing on large amounts of real-time heterogeneous data from sensor and social media. In addition to research, he enjoys teaching Machine Learning as Adjunct Faculty member in the School of Computing at the Dublin City University. Jia Yuan completed his PhD in electrical engineering at McGill University (2010), and has worked at Intel Research, Stanford University, and Ecole Normale Superieure Paris.