Damian Niwinski

Information theory

Lecture: Tuesday 10:15-12, room 3160.
Lecture Notes On-Line ps, pdf, updated 17.05.2006.
Comments welcome!

Tutorials: Tuesday 12:15-14, room 2100, Michal Strojnowski.
Friday 8:30-10, room 1030, Hugo Gimbert.
Tutorials web page

Exam Zadania zaliczeniowe new series!

Objectives:
Introduction into a theory which is useful in many application of informatics, like cryptography, modeling of natural language, and bio-informatics. The theory defines quantitative measures of information contained in a random variable or in a sequence of bits.
It also provides criteria of optimal compressing (coding) of information, and of sending a message through an insecure channel.

Plan:
1. From the 20 questions game to the concept of entropy. Kraft inequality. Codes of Huffman and Shannon-Fano.
2. Conditional entropy, information.
3. The First Shannon Theorem about optimal encoding.
4. Channels, information lost, improving efficiency, channel capacity.
5. The Shannon Theorem about sending information through a noisy channel.
6. Information complexity by Kolmogorov. Chaitin number.
7. Kolmogorov's complexity vs. Shannon's entropy---universal test by Martin Lof.

Literature (suggestions welcome!).

Previous edition: Teoria informacji 2004.