Imperial College London > Talks@ee.imperial > Featured talks > Encoding Tasks Revisited: Sources with Memory, Mismatch, and a Divergence

Encoding Tasks Revisited: Sources with Memory, Mismatch, and a Divergence

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Patrick Kelly.

In the task-encoding problem (a.k.a. Honey-Do problem) a source sequence is described using a fixed number of bits, based on which a list that is guaranteed to contain the source sequence must be produced. We seek the shortest description length that allows some moment of the list size to approach one.

This problem was recently solved for memoryless sources using a technique that is based on the method of types, but the technique does not extend to general sources with memory.

In this talk I shall present a different approach that works also for sources with memory, provided that their Renyi-entropy rate is well defined. The new approach enables us to study the penalty for designing the scheme based on the wrong source law (mismatch) and, in this way, motivates a new definition of Renyi Divergence.

Based on joint work with Christoph Bunte.

This talk is part of the Featured talks series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

Changes to Talks@imperial | Privacy and Publicity