Latest News and Events

The SAMSI-FODAVA Workshop on Interactive Visualization and Analysis of Massive Data will be held on December 10-12, 2012.
Posted: October 02, 2012
The FODAVA Annual Meeting will immediately follow (Dec 12-13) the SAMSI/FODAVA joint workshop at the same location.
Posted: September 05, 2012
Many of the modern data sets such as text and image data can be represented in high-dimensional vector spaces and have benefited from computational methods that utilize advanced techniques from num
Posted: June 30, 2012

Sparse Recovery Problems in Machine Learning

Speaker: 
Vladimir Koltchinskii
Abstract: 

Several methods of sparse recovery in empirical risk minimization
problems will be discussed. These methods are based on special
penalization techniques such as LASSO or l_1 penalization.
They can be used in a variety of problems of Machine Learning
such as regression and pattern classification and they provide
a way to find a sparse approximation of an unknown target function
and to reduce the dimensionality of the data. Recent attempts to
understand these problems mathematically are based on methods of asymptotic
geometric analysis and high dimensional probability.
The development of a theory and methods of sparse recovery
beyond the usual framework of finite dictionary to more
general settings of learning theory is a challenging open problem.