13
Intelligent Database Systems Presenter : Fen-Rou Ciou Authors : M.H. Ghaseminezhad, A. Karami 2011,ASC A novel self-organizing map (SOM) neural network for discrete groups of data clustering

Presenter : Fen-Rou Ciou Authors : M.H. Ghaseminezhad , A. Karami 2011,ASC

  • Upload
    ohio

  • View
    44

  • Download
    0

Embed Size (px)

DESCRIPTION

A novel self-organizing map (SOM) neural network for discrete groups of data clustering. Presenter : Fen-Rou Ciou Authors : M.H. Ghaseminezhad , A. Karami 2011,ASC. Outlines. Motivation Objectives Methodology Experiments Conclusions Comments. Motivation. - PowerPoint PPT Presentation

Citation preview

Page 1: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Presenter : Fen-Rou Ciou

Authors : M.H. Ghaseminezhad, A. Karami

2011,ASC

A novel self-organizing map (SOM) neural network for discrete groups of data

clustering

Page 2: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

OutlinesMotivationObjectivesMethodologyExperimentsConclusionsComments

Page 3: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Motivation• However, no algorithm that can automatically

cluster discrete groups of data is presented, and

our simulation results show that the classic SOM

algorithm cannot cluster discrete data correctly.

Page 4: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Objectives• In this paper present a novel SOM-based algorithm

that can automatically cluster discrete groups of data

using an unsupervised method.

Page 5: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Methodology

Page 6: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Methodology – First PhaseInitialize

Find the value

Choose an input vector X

Calculate the distance

Update weights

Set t = t+1Until t = T

Initialize all weights wij

Set Learning iteration number t=0, topological neighborhood d0 , Learning rate ᾳ0 , Total iterations T , Total number of neurons M

t < T

Page 7: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Methodology – Second PhaseInitialize

Calculate “batch”

parameter

Find the value

Choose an input vector X

Calculate the distance

Increase the age

Update weights

If b < Batch

Update weights

Set t = t+1Until t = T

Initialize all weights wij

Set Learning iteration number t=0, topological neighborhood d0 , Learning rate ᾳ0 , Total iterations T , Total number of neurons MBatch0 ,

Page 8: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Methodology – Third PhaseInitial M and set i

= 1Choose

Union also set i

= i + 1If i < M

Page 9: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Methodology

Page 10: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Experiments

Page 11: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Experiments

Page 12: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Conclusions• The novel SOM algorithm does a substantially better

job of clustering discontinues data as a result of its

flexible structure as well as employing the batch

learning method.

Page 13: Presenter   :  Fen-Rou  Ciou Authors      : M.H.  Ghaseminezhad , A.  Karami 2011,ASC

Intelligent Database Systems Lab

Comments• Advantages

• Applications– SOM