Presented By O'Reilly and Cloudera
Make Data Work
Sept 29–Oct 1, 2015 • New York, NY

Tensor methods: A new paradigm for training probabilistic models and for feature learning

Anima Anandkumar (UC Irvine)
10:00am–10:30am Tuesday, 09/29/2015
Hardcore Data Science
Location: 1 E10/1 E11 Level: Advanced
Average rating: **...
(2.50, 4 ratings)

Tensors are rich structures for modeling complex higher order relationships in data-rich domains such as social networks, computer vision, and the Internet of Things. Tensor decomposition methods are embarrassingly parallel and scalable to enormous datasets. They are guaranteed to converge to the global optimum and yield consistent estimates of parameters for many probabilistic models such as topic models, community models, and hidden Markov models.

I will show the results of these methods for learning topics from text data, communities in social networks, disease hierarchies from healthcare records, cell types from mouse brain data, etc. I will also demonstrate how tensor methods can yield rich discriminative features for classification tasks and serve as an alternative method for training neural networks.

Photo of Anima Anandkumar

Anima Anandkumar

UC Irvine

Anima Anandkumar is a principal scientist at Amazon Web Services. Anima is currently on leave from UC Irvine, where she is an associate professor. Her research interests are in the areas of large-scale machine learning, nonconvex optimization, and high-dimensional statistics. In particular, she has been spearheading the development and analysis of tensor algorithms. Previously, she was a postdoctoral researcher at MIT and a visiting researcher at Microsoft Research New England. Anima is the recipient of several awards, including the Alfred. P. Sloan fellowship, the Microsoft faculty fellowship, the Google research award, the ARO and AFOSR Young Investigator awards, the NSF CAREER Award, the Early Career Excellence in Research Award at UCI, the Best Thesis Award from the ACM SIGMETRICS society, the 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, and O’Reilly Media. Anima holds a BTech in electrical engineering from IIT Madras and a PhD from Cornell University.

Comments on this page are now closed.

Comments

Robert Lucente
09/02/2015 6:51am EDT

Thanks for the heads up about your presentation. Look forward to it.

Picture of Anima Anandkumar
Anima Anandkumar
08/15/2015 10:49am EDT

I plan to have a substantially different talk and show how we can use tensor methods for training neural networks and for feature learning

Robert Lucente
08/08/2015 7:49am EDT

I watched the presentation that was given at Strata at San Jose at YouTube at

https://www.youtube.com/watch?v=KmvZu9qJNzg

Impressive. I was wondering how the presentation in NYC will be different?