Speaker:
Date of Talk:
12/02/10
Affiliation:
Department of Applied Mathematics, University of Colorado at Boulder
Title:
Uncovering Local Manifold Geometry and Processing Large Data Sets
Abstract
The volume of data created by science and society is growing at an unprecedented rate. As our ability to collect and store information is ever growing, it is essential that our ability to understand and thus process these massively large data sets keeps pace.
Many data sets, while presented in high dimension, are actually organized along a lower dimensional manifold. Processing such data sets then becomes a problem of learning the geometry of this manifold. Traditional algorithms proceed by producing a global representation of the data. We are developing techniques for learning a manifold's geometric properties at a local scale and are designing algorithms that exploit this local information to allow for processing (e.g., approximation, denoising, searching) of data to an accuracy limited only by the original sampling.
