Compressive Learning: developing and analyzing new algorithms for compressive learning and investigating their computational and statistical efficiency in processing big data

A classical problem in data analysis is how to compress the size of data without destroying its salient information. This is becoming increasingly important in the era of big data where we need to process ever increasing quantities of data of various modalities. Some exciting new developments in this regard have recently emerged exploiting ideas such as kernel representations [4] and ‘sketching’ [5] in new "compressive learning" algorithms [1, 2, 3].

This project will develop and analyze new algorithms for compressive learning and investigate their computational and statistical efficiency in processing big data. The studentship forms a key component of the advanced ERC project C-SENSE, "Exploiting Low Dimensional Models in Sensing, Computation and Processing".

The successful candidate will be located in the Prof Davies’ research group in the Institute for Digital Communications (IDCOM). There will also be opportunities to collaborate with the University of Edinburgh’s School of Mathematics and the Alan Turing Institute for Data Science.

TO APPLY - please go to the University’s Jobs page and search for vacancy 040772.

[1] N. Keriven, A. Bourrier, R. Gribonval, P. Perez. Sketching for Large Scale Learning of Mixture Models. ICASSP, 2016.
[2] A. Rahimi and B. Recht, Random Features for Large Scale Kernel Machines. Advances in Neural Information Processing Systems, no. 1, 2007.
[3] H. Reboredo, F. Renna, R. Calderbank, and M. D. Rodrigues, Compressive Classification. Proc. GlobalSIP, no. 1, 2013.
[4] B. Scholkopf and A. Smola, Learning with Kernels. MIT Press, Cambridge, MA, 2002.
[5] D. Woodruff, Sketching as a Tool for Numerical Linear Algebra. Foundations and Trends in Theoretical Computer Science Vol. 10, No. 1-2 (2014).