6.441 Information Theory
Mathematical definitions of information measures, convexity, continuity, and variational properties. Lossless source coding; variable-length and block compression; Slepian-Wolf theorem; ergodic sources and Shannon-McMillan theorem. Hypothesis testing, large deviations and I-projection. Fundamental limits of block coding for noisy channels: capacity, dispersion, finite blocklength bounds. Coding with feedback. Joint source-channel problem. Rate-distortion theory, vector quantizers. Advanced topics include Gelfand-Pinsker problem, multiple access channels, broadcast channels (depending on available time).
This class has 6.041B as a prerequisite.
Lecture occurs 11:00 AM to 12:30 PM on Tuesdays and Thursdays in 4-149.
This class counts for a total of 12 credits.
© Copyright 2015 Yasyf Mohamedali