CYUT Information Management Inference of Recommendation Information on the Internet Using Improved FAM Authors: Won Kim, Il-Ju Ko, Jin-Sung Yoon, Gye- Young Kim Source: Future Generation Computer Systems, Vol. 20, No.2, pp. 265-273, Feb. 2004 Instructor Professor: Dr. Rong-Chang Chen Presented By: Yu-Chun Chen ( 陳陳陳 ) Wan-Jing Liu ( 陳陳陳 ) Date: 2005/01/12
20
Embed
CYUT Information Management Inference of Recommendation Information on the Internet Using Improved FAM Authors:Won Kim, Il-Ju Ko, Jin-Sung Yoon, Gye-Young.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CYUT Information Management
Inference of Recommendation Information on the Internet Using
Improved FAM
Authors: Won Kim, Il-Ju Ko, Jin-Sung Yoon, Gye-Young Kim
CYUT Information ManagementFuzzy Associative Memory (3/3)
Introduction ExperimentalIFAM ConclusionComment
input association
output
0.8
0.5
0.6
0.3
0.2
0.6
0.3
0.4
0.3
2005/01/12 10
CYUT Information Management
Inference model with IFAMIntroduction ExperimentalIFAM ConclusionComment
2005/01/12 11
CYUT Information Management
Membership Function
Divide the entire range into three parts and assign then with low, medium, high fuzzy sets
0.1
0.8 low
0.7
0.4 high
0.33 medium
Introduction ExperimentalIFAM ConclusionComment
2005/01/12 12
CYUT Information Management
D value
The degree of usefulness of input fuzzy setsM : the number of recommendation itemsK : the total number of learning dataPAl(j) : the probability density of rates on the j th recommendation itemrBj(xi) : the rate of xi on a recommendation itemAi(xi) : the fit value of xi to an input fuzzy set Ai
Introduction ExperimentalIFAM ConclusionComment
2005/01/12 13
CYUT Information Management
W value
Hebbian-style learning method η: positive learning rateai (n) : input associate for the n th learning databj (n) : output associate for the n th learning data