Huffman Code

  • Entropy: define the minimum of expected code length.

  • high entropy -> uniform tree, unpredictable; low entropy -> skewed tree, predictable

  • Heap: complete binary tree. Max-heap, parents always bigger than children.

  • time complexity of insert() in heap: O(logN), bubble up the child node

  • time complexity of remove() in heap: O(logN), the last node move to root, then trickle down to the right place.

  • prefix code: no symbol’s codeword is a prefix of another

REFERENCE: UCSD CSE100 Podcast 2/22/2017

results matching ""

    No results matching ""