|
 |
|
|
Thursday, February 09, 2012
|
Sofus Macskassy: Studies of Inference in Networks |
Event Type: CS Colloquium
Time: 3:30 PM - 5 PM
Location: SSL 150
Title:
Studies of Inference in Networks: Classification, Combining and Creating Networks for better Classification
Abstract:
There has been an increasing amount interest and published work in classification of relational data over the past decade. In fact, available relational data in various formats are growing in size and complexity, and we need techniques for analyzing these data. I will
in this talk start with describing a network-based collective inference technique for classifying nodes in a large partially-labeled network. This is also known as within-network classification and is analogous to transductive and semi-supervised learning in standard machine learning. I will continue to discuss how we use this technique in the case where we have different type of data about the same nodes (e.g., links and text). Specifically, I will show how we can objective combine different types of link. Finally, I will discuss how we can in fact take non-relational data and create a network in a principled manner such that we can leverage the power of network-based inference to improve classification over standard machine learning.
Bio:
Sofus A. Macskassy is a Sr. Computer Scientist at the Information Sciences Institute at the University of Southern California. He was previously the Director of Fetch Labs at Fetch Technologies. His work spans a wide area of research and domains, including social media analytics, where he is extracting and analyzing user profiles and behaviors from Twitter. He is also applying statistical relational learning techniques to study classification of relational data. For example, he has studied how to combine text and links for better classification, identification of suspicious entities in social networks, among others. He is the primary developer and maintainer of the open-source Network Learning Toolkit for Statistical Relational Learning. He is published in major venues, has served on organizing and programming committees in major conferences. He serves on the editorial board of the Machine Learning Journal, the premier machine learning journal in the world. |
|
Tuesday, February 14, 2012
|
Dr. Jennifer Chayes, Microsoft Research New England |
Event Type: Distinguished Lecture
Time: 3:30 PM - 5 PM
Location: SSL 150
Speaker: Dr. Jennifer Chayes
Distinguished Scientist and Managing Director, Microsoft Research New England
Abstract:
Everywhere we turn, we find that networks have become increasing appropriate descriptions of relevant interactions. In the high tech world, we see mobile networks, the Internet, the World Wide Web, and a variety of online social networks. In economics, we are increasingly experiencing both the positive and negative effects of a global networked economy. In epidemiology, we find disease spreading over our ever growing social networks, complicated by mutation of the disease agents. In problems of world health, distribution of limited resources, such as water resources, quickly becomes a problem of finding the optimal network for resource allocation. In biomedical research, we are beginning to understand the structure of gene regulatory networks, with the prospect of using this understanding to manage the many diseases caused by gene misregulation. In this talk, I look quite generally at some of the models we are using to describe these networks, and at some of the methods we are developing to indirectly infer network structure from measured data. In particular, I will discuss models and techniques which cut across many disciplinary boundaries.
Bio:
Jennifer Tour Chayes is Distinguished Scientist and Managing Director of Microsoft Research New England in Cambridge, Massachusetts, which she co-founded in July 2008. Before this, she was Research Area Manager for Mathematics, Theoretical Computer Science and Cryptography at Microsoft Research Redmond. Chayes was also the former Vice-President of the American Mathematical Society and was for many years Professor of Mathematics at UCLA until 1997, when she joined Microsoft Research and co-founded the Theory Group. She received her Ph.D. in mathematical physics at Princeton and B.A. in biology and physics at Wesleyan University.
Her research areas include phase transitions in discrete mathematics and computer science, structural and dynamical properties of self-engineered networks, and algorithmic game theory. She is the co-author of over 110 scientific papers and the co-inventor of more than 25 patents. Chayes is well known for her work on phase transitions, in particular for laying the foundation for the study of phase transitions in problems in discrete mathematics and theoretical computer science; this study is now giving rise to some of the fastest known algorithms for fundamental problems in combinatorial optimization. Among Chayes' contributions to Microsoft technologies are the development of methods to analyze the structure and behavior of various networks, the design of auction algorithms, and the design and analysis of various business models for the online world. She is the recipient of a Sloan Fellowship and the UCLA Distinguished Teaching Award, and a Fellow of ACM, AAAS, the Fields Institute. She is also currently a member of the ACM Turing Award Selection Committee.
Chayes lives with her husband, Christian Borgs, who also happens to be her principal scientific collaborator. In her spare time, she enjoys overworking. |
|
| To add events here please contact
us at: |
csciasst |
@usc.edu |
|
|
|
|
