Wednesday, May 22, 2013

Slides from the workshop on Big Data: Theoretical and Practical Challenges

May 148h30 - 9h10 : Registration and coffee
9h10 - 9h20 : Introduction
9h20 - 10h10 : Chris Holmes, Oxford University
, Bayesian Hidden Markov models with linear time decoding for the analysis of cancer genomes10h10 - 10h50 : Coffee break
10h50 - 11h40 : Eric Moulines, Telecom Paristech, Islands Particle model11h40 - 12h30 : Sonia Petrone, Università Bocconi
, Restricted random partitions for Bayesian curve fitting12h30 - 14h : Lunch Buffet
14h - 14h50 : Michael Jordan, U.C. Berkeley
, MAD-Bayes: MAP-based asymptotic derivations from Bayes14h50 - 15h40: Alexandre d'Aspremont, CNRS - Ecole Polytechnique
, Approximation Bounds for Sparse Principal Component Analysis15h40 - 16h20: Coffee break
16h20 - 17h10 : Alfred Hero, University of Michigan, 
Correlation mining17h10 - 18h: Martin Wainwright, U.C. Berkeley, 
Computation meets Statistics: Fast global convergence for high-dimensional (non-convex) statistical recovery
May 159h10 - 10h : Leon Bottou, Microsoft Research, 
Large-Scale Learning Revisited
10h - 10h40 : Coffee break
10h40 - 11h30 : Francis Bach, INRIA - ENS
, Stochastic gradient methods for large-scale machine learning11h30 - 12h20 : Ion Stoica, U.C. Berkeley, 
Computations with Bounded Errors and Bounded Response Times on Very Large Data12h20 - 14h : Lunch (take-out)
14h - 14h50 : Piotr Indyk, 
MIT, Faster Algorithms for the Sparse Fourier Transform
14h50 - 15h40:  Slav Petrov, Google, Large-Scale Language Learning
 15h40 - 16h20: Coffee break
16h20 - 17h10 : Lester Mackey, Stanford University
, Divide-and-Conquer Matrix Factorization17h10 - 18h: Michael Mahoney, Stanford University
, Revisiting the Nystrom Method for Improved Large-Scale Machine Learning18h - 18h20: Conclusion

No comments: