Guaranteed Non-convex Machine Learning Algorithms through Spectral Methods
- 14:00 2nd December 2016 ( week 8, Michaelmas Term 2016 )Tony Hoare Room, Robert Hooke Building
Most learning problems can be cast as optimization tasks which are non-convex. Developing fast and guaranteed approaches for solving non-convex problems is a grand challenge. I will show how spectral optimization can reach the globally optimal solution for many learning problems despite being non-convex. This includes unsupervised learning of latent variable models, training neural networks and reinforcement learning of partially observable Markov decision processes. It involves spectral decomposition of moment matrices and tensors. Tensors are rich structures that can encode higher order relationships in data. In practice, tensor methods yield enormous gains both in running times and learning accuracy over traditional methods such as variational inference. Overall, these positive results demonstrate that many challenging learning tasks can be solved with guarantees using efficient non-convex approaches. I will end the talk by mentioning my ongoing work at Amazon on large-scale learning.
Speaker bio
Anima Anandkumar is a principal scientist at Amazon Web Services, and is currently on leave from U.C.Irvine, where she is an associate professor. Her research interests are in the areas of large-scale machine learning, non-convex optimization and high-dimensional statistics. In particular, she has been spearheading the development and analysis of tensor algorithms. She is the recipient of several awards such as the Alfred. P. Sloan Fellowship, Microsoft Faculty Fellowship, Google research award, ARO and AFOSR Young Investigator Awards, NSF CAREER Award, Early Career Excellence in Research Award at UCI, Best Thesis Award from the ACM SIGMETRICS society, IBM Fran Allen PhD fellowship, and several best paper awards. She has been featured in a number of forums such as the Quora ML session, Huffington post, Forbes, O’Reilly media, and so on. She received her B.Tech in Electrical Engineering from IIT Madras in 2004 and her PhD from Cornell University in 2009. She was a postdoctoral researcher at MIT from 2009 to 2010, an assistant professor at U.C. Irvine between 2010 and 2016, and a visiting researcher at Microsoft Research New England in 2012 and 2014Share this:
