Poster
Estimation, Optimization, and Parallelism when Data is Sparse
John Duchi · Michael Jordan · Brendan McMahan
Harrah's Special Events Center, 2nd Floor
We study stochastic optimization problems when the \emph{data} is sparse, which is in a sense dual to the current understanding of high-dimensional statistical learning and optimization. We highlight both the difficulties---in terms of increased sample complexity that sparse data necessitates---and the potential benefits, in terms of allowing parallelism and asynchrony in the design of algorithms. Concretely, we derive matching upper and lower bounds on the minimax rate for optimization and learning with sparse data, and we exhibit algorithms achieving these rates. Our algorithms are adaptive: they achieve the best possible rate for the data observed. We also show how leveraging sparsity leads to (still minimax optimal) parallel and asynchronous algorithms, providing experimental evidence complementing our theoretical results on medium to large-scale learning tasks.
Live content is unavailable. Log in and register to view live content