This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Hu — Tucker
__’____________——————— a Anode isaEloraO
a Two nodes are compatible if there is no E] between them
11 A weight sequence is a sequence of nodes with weight
:1 Weight of a father is the sum of the two sons' weights
1: Position of a father is the position of the left son
a Given two adjacent nodes of equal weights, the left is considered less 1/11 Hu  Tucker Step I: Combination Combine the min weight compatible pair until
there is a tree T’ 2/11 Hu — Tucker Step II: Get levels of D in T’ o‘@ 3/11 FhJ— Thcker Step III: Reconstruction 4/n Hu — Tucker ________—_—_____——————————— Read the levels from left to right. Use a stack. 2, 3,4,4, 2, 2
3
2 2 MOOF34:
MUJUJ
[\JM —ll\.)l\) 21
1110 lMUOP 5/11 HuThcker Step I: Combination Instead of min weight compatible pair, use a
local min weight c0mpatible pair in a weight
sequence. @El,
IfEl>andlE<JhenIEandis a local minimum.  6/11 7/11 Hu — Tucker @
@ Two circle at the bottom
One circle and one square Two squares on the slope 8/11 Hu  Tucker __________—______——————————— For an increasing weight sequence, Huffman
: Hu—Tucker, so the tree T’ is optimum. For a valley sequence, the tree T’ is optimum. E] Q E]
DOE]
0 9/11 Hu — Tucker T’ can be normalized. 10/11 Hu — Tucker 11/11 ...
View
Full Document
 Fall '06
 Hu
 Algorithms

Click to edit the document details