Date: Wednesday, 9 January 2008, 6:30 PM
Location: SAP LABS, Building D, 3410 Hillview Avenue,
Palo Alto, CA (Google
Maps | Yahoo!
Maps | Mapquest)
Cost: Free and open to all who wish to attend, but membership is
only $10/year.
Topic
Hebbian learning is a very know principal of unsupervised learning in networks: if two events happen "close in time" then the strength of connection between the network nodes producing those events increases. Is it a complete set of learning axioms? Given a reinforcement signal (reward) for a sequence of actions we can add another axiom: reward controls plasticity. Thus, we get a reinforcement learning algorithm that could be used for training spiking neural networks. The author will demonstrate the utility of this algorithm on a maze learning problem. Can these learning principles be applied not only to neural, but also to other kinds of networks? Yes, in fact we will see their application to economical influence networks for portfolio optimization. Then (if time allows) we consider another application: social networks for a movie recommendation engine, and other causality inducing principles instead of "close in time". By the end of the talk the author hopes that the audience would agree that the reward controls plasticity principle is a vital learning axiom.
About the Speaker
Victor V. Miagkikh received his Engineering Diploma degree in Computer Engineering from Taganrog State University of Radio Engineering, Russia in 1994. He revived his M.S. in Computer Science from Michigan State University in 1998. He is a Ph.D. candidate in Computer Science at Michigan State University, and works on a certificate in data mining at Stanford University as of today. He is a Machine Learning Specialist with Cisco Systems, Inc. (Ironport business unit). His current research is focusing on spam detection and reputation filtering algorithms. Victor also worked in research and development of Procter & Gamble, Inc. and Fair Isaac, Inc. Among his awards: Presidential Scholarship for Study Abroad, and King-Size Red Diploma (the best graduate of the department). His research interest are machine learning, artificial intelligence, spam detection, reinforcement learning, neural networks, genetic algorithms, fuzzy logic, social networks, reputation systems, multi aspect graph theoretic modeling, collaborative filtering, network analysis, econometric and psychometric modeling. His publications could be found here: http://www.msu.edu/~miagkikh/
Video
Slides from Presentation [Powerpoint]
Slides from Presentation [PDF]