CS Department Seminar: Dr. John Langford, Microsoft Research NY

Title: Extreme Multiclass

Wednesday, November 13, 2013 ( 11:00 am to 12:00 pm )

Location: Babbio Center 310

Abstract:

Standard multiclass classification techniques require O(k) computation during train and test where k is the number of classes, yet the information theoretic lower bound is O(log k). This gap matters little when k is on the order of 10 or 100, but at 10^4  or 10^6  it matters a great deal. I will discuss the theory of extreme multiclass classification including consistency, robustness, efficiency, and structure learning.

Bio:

John Langford is a machine learning research scientist, a field which he says "is shifting from an academic discipline to an industrial tool". He is the author of the weblog hunch.net and the principal developer of Vowpal Wabbit.  John works at Microsoft Research New York, of which he was one of the founding members, and was previously affiliated with Yahoo! Research, Toyota Technological Institute, and IBM's Watson Research Center. He studied Physics and Computer Science at the California Institute of Technology, earning a double bachelor's degree in 1997, and received his Ph.D.in Computer Science from Carnegie Mellon University in 2002. He was the program co-chair for the 2012 International Conference on Machine Learning.

 

For additional information please contact:
Jingrui He
Assistant Professor
Lieb 311
Phone: 201-216-5358
jingrui.he@cs.stevens.edu