Error exponents for source coding under logarithmic loss

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

• Dr Hamdi Joudeh
• Wednesday 03 May 2023, 14:00-15:00
• MR5, CMS Pavilion A.

If you have a question about this talk, please contact Dr Varun Jog.

In this talk, I will discuss the problem of lossy source coding under the logarithmic loss distortion measure. This distortion measure is a natural choice in settings where the decoder (or decompressor) produces a soft reconstruction of the source, i.e., a probability distribution, instead of a point estimate. I will focus on block-coding in discrete memoryless settings, and present new results on the error exponent under an excess distortion constraint. I will then extend the results to settings with side information. The proofs are based on the method of types; and make use of the close relationship between lossy source coding under logarithmic loss, and almost lossless source coding with list decoding.

This talk is part of the Information Theory Seminar series.

This talk is included in these lists:

• All CMS events
• All Talks (aka the CURE list)
• CMS Events
• DPMMS Lists
• DPMMS info aggregator
• DPMMS lists
• Hanchen DaDaDash
• Information Theory Seminar
• Interested Talks
• MR5, CMS Pavilion A
• School of Physical Sciences
• Statistical Laboratory info aggregator
• bld31

Note that ex-directory lists are not shown.