Conference abstracts

Session A1 - Approximation Theory

July 10, 17:00 ~ 17:35

Tensorization for data-sparse approximation

Lars Grasedyck

RWTH Aachen University, Germany   -   lgr@igpm.rwth-aachen.de

In this talk we will review techniques for the tensorization of functions or vectors that are given as low dimensional objects. The tensorization artificially casts them into high dimensional objects where we are able to apply low rank tensor approximation techniques. The rationale behind this is the fact that low rank tensors can be represented with a complexity linear in the dimension, thus allowing a logarithmic complexity for the representation of objects that --- when stored in a naive way --- would require a linear complexity. We characterize the approximability in terms of subspaces and give examples for classes of functions that allow for such a compression of data.

View abstract PDF



FoCM 2017, based on a nodethirtythree design.