Can we optimize data by fixing the encoding length instead of the message?
In this lecture from Sam Cohen's third-year Information Theory course at Oxford, we explore Tunstall’s algorithm. Discover how to build optimal codes by varying input length while maintaining a fixed encoding length.
While standard variable-length encodings focus on compressing messages of different lengths, Tunstall’s algorithm approaches the problem from the opposite direction. It considers a system where the encoding length remains fixed, but the length of the input messages varies.
This mathematical approach provides an optimal method for constructing such codes, offering a unique perspective on the fundamental mechanics of information theory.
Source: Information Theory, Lecture 8: Tunstall’s code - Oxford Mathematics 3rd Year Student Lecture
