Alternate Learning and Compression approaching R(D)

Ram Zamir, Kenneth Rose

Keywords: rate-distortion, on-line learning, exploration exploitation, Blahut algorithm

TL;DR: Adaptive rate-distortion in view of on-line learning, exploration-exploitation tradeoff, and "natural type selection".

Abstract:

The inherent trade-off in on-line learning is between exploration and exploitation.
A good balance between these two (conflicting) goals can achieve a better
long-term performance. Can we define an optimal balance? We propose to study
this question through an adaptive lossy compression system, which exhibits a
”natural” trade-off between exploration and exploitation.

Submission Number: 17