Tradeoffs in Large Scale Learning: Statistical Accuracy vs. Numerical Precision

Channel:
United States Simons Institute for the Theory of Computing
Subscribers:
68,700
Published on ● Video Link: https://www.youtube.com/watch?v=Bjjj8VYiKsQ


Duration: 32:49
605 views
8

Sham Kakade, Microsoft Research New England
Information Theory, Learning and Big Data
http://simons.berkeley.edu/talks/sham-kakade-2015-03-19



Other Videos By Simons Institute for the Theory of Computing


2015-03-25Machine Learning from Human Comparative Judgments
2015-03-25Minimum Rényi Correlation Principle: From Marginals to Joint Distribution
2015-03-25Near-Optimal Hypothesis Testing via Convex Optimization
2015-03-25Sketching for M-Estimators: A Unified Approach to Robust Regression
2015-03-25Data Driven Convergence in Statistical Estimators
2015-03-25Learning a Hidden Basis through Imperfect Measurements: Why and How
2015-03-25Efficient Minimax Optimal Strategies for Universal Prediction
2015-03-25Simple, Efficient and Neural Algorithms for Sparse Coding
2015-03-25Lucky Talk: Chebyshev Polynomials, Moment Matching and Optimal Estimation of the Unseen
2015-03-25Information Measure Estimation and Applications: Boosting the Sample Size from n to n log n
2015-03-25Tradeoffs in Large Scale Learning: Statistical Accuracy vs. Numerical Precision
2015-03-25Learning and Testing Structured Distributions
2015-03-25Statistical Property Testing and Estimation Beyond the i.i.d. Setting
2015-03-25Computation-Statistics Tradeoffs in Unsupervised Learning via Data Summarization
2015-03-25On Learning Distributions from their Samples
2015-03-25Confidence-Based Active Learning
2015-03-25Lucky Talk: Low-regret Recommendations and Item-item Collaborative Filtering
2015-03-25Learning Sparse Data with Near-optimal Speed and Efficiency from a Variety of Measurement Processes
2015-03-25Approximating Spherical Gaussian Mixtures
2015-03-25Constructing Informative Features for Discriminative Learning
2015-03-25Strong Data Processing Inequalities: Applications to MCMC and Graphical Models


Tags:
Simons Institute
UC Berkeley
computer science
theory of computing
Information Theory
Sham Kakade