Download presentation
Presentation is loading. Please wait.
Published byElinor Cole Modified over 9 years ago
1
Intelligent Database Systems Lab Presenter : Chang,Chun-Chih Authors : Miin-Shen Yang a*, Wen-Liang Hung b, De-Hua Chen a 2012, FSS Self-organizing map for symbolic data
2
Intelligent Database Systems Lab Outlines Motivation Objectives Methodology Experiments Conclusions Comments
3
Intelligent Database Systems Lab Motivation SOM neural network is constructed as a learning algorithm for numeric (vector) data. There is less consideration in a SOM clustering for symbolic data.
4
Intelligent Database Systems Lab Objectives We then use a suppression concept to create a learning rule for neurons. The S-SOM is created for treating symbolic data by embedding the novel structure and the suppression learning rule. This paper can treat symbolic data and a so-called symbolic SOM (S-SOM) is then proposed.
5
Intelligent Database Systems Lab Methodology SOM for numeric data
6
Intelligent Database Systems Lab Methodology Quantitative type of A k and B k
7
Intelligent Database Systems Lab Methodology Qualitative type of A k and B k
8
Intelligent Database Systems Lab Methodology calculate the dissimilarity measure between object 1 and 10
9
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
10
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
11
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
12
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
13
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
14
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
15
Intelligent Database Systems Lab Methodology Calculate the degree of membership Measure Xi and Nj distance Calculatin g the hj(t) Calculating the learning rate
16
Intelligent Database Systems Lab Experiments
17
Intelligent Database Systems Lab Experiments
18
Intelligent Database Systems Lab Experiments
19
Intelligent Database Systems Lab Experiments
20
Intelligent Database Systems Lab Experiments
21
Intelligent Database Systems Lab Experiments
22
Intelligent Database Systems Lab Experiments
23
Intelligent Database Systems Lab Experiments
24
Intelligent Database Systems Lab Experiments - Clustering result from our method
25
Intelligent Database Systems Lab Experiments -Clustering result of IFCM
26
Intelligent Database Systems Lab Experiments -Clustering result from our method
27
Intelligent Database Systems Lab Experiments -37 countries every month temperature
28
Intelligent Database Systems Lab Experiments 5.Cairo 開羅 19. Mauritius 摩里斯理 7.Colombo 巴拉那州
29
Intelligent Database Systems Lab Conclusions The S-SOM can be effective in clustering and also responds information of input symbolic data. The experimental results also demonstrated that the S- SOM is feasible to treat symbolic data.
30
Intelligent Database Systems Lab Comments Advantages - The experimental results also demonstrated that the S-SOM is feasible to treat symbolic data. Applications - Self-organizing map of Symbolic data
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.