[Ml-stat-talks] William Pierson Field Lecture on Kernels and Deep Learning (Today, 1:30 PM, Bendheim 103)
Amir Ali Ahmadi
a_a_a at princeton.edu
Tue Apr 26 01:47:36 EDT 2016
The William Pierson Field Lecture below is primarily meant for students of ORF 523, but all interested are welcome to attend.
DATE: Today (Tuesday, April 26)
TIME: 1:30 PM - 2:50 PM
LOCATION: Bendheim Center for Finance, Room 103
SPEAKER: Vikas Sindhwani, Google Research
TITLE: Kernels, Random Embeddings and Deep Learning
Kernel methods and Neural Networks are two great algorithmic traditions in machine learning for estimating non-linear dependencies in complex datasets. In recent years, powered by large datasets and compute resources, Deep Neural Networks have led to dramatic improvements in various domains; they are widely deployed across Google for enabling speech recognition, computer vision and natural language processing applications. In the mid-nineties, Kernel methods emerged as a methodology of choice due to their elegant mathematical underpinnings and convex optimization formulations. In this lecture, I'll share some personal perspectives around these traditions, give an overview of key ideas in kernel methods including a randomization technique to alleviate their scalability challenges, and discuss some connections, challenges and open problems.
Vikas Sindhwani is Research Scientist in the Google Brain team in New York. His interests are broadly in core mathematical foundations of statistical learning, and in end-to-end design aspects of building large-scale, robust machine intelligence systems. He received the best paper award at Uncertainty in Artificial Intelligence (UAI) 2013, the IBM Pat Goldberg Memorial Award in 2014, and was co-winner of the Knowledge Discovery and Data Mining (KDD) Cup in 2009. He previously led the Machine Learning group at IBM Research, NY, and has a PhD in CS from the University of Chicago. His publications are available at: http://vikas.sindhwani.org/.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Ml-stat-talks