References in periodicals archive ?
The new approach we present is also information theoretic and uses tools from the theory of Huffman codes to develop a deterministic construction of a sequence of binary sampling vectors [a.
This should not be a surprise since the algorithm was inspired by the theory of Huffman codes.
Once the Huffman code table of each letters is generated, we compress each verse by writing the code of its characters in the compressed file.
While Huffman has already known that the average code length is asymptotically equal to the entropy of the source, the asymptotic performance of the Huffman code is still not fully understood.