self organizing maps n.
Download
Skip this Video
Loading SlideShow in 5 Seconds..
Self-Organizing Maps PowerPoint Presentation
Download Presentation
Self-Organizing Maps

Loading in 2 Seconds...

play fullscreen
1 / 16

Self-Organizing Maps - PowerPoint PPT Presentation


  • 179 Views
  • Uploaded on

Self-Organizing Maps. Cao Mai December 14, 2009. Outline. Define Self-Organizing Maps (SOM) Quick Example Explain SOM An example of SOM using MGP dataset Questions. Define. A Self-Organizing Map (SOM) is a way to visually represent multi-dimensional data in an usually 2-D or 3-D manner

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about 'Self-Organizing Maps' - chalice


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
self organizing maps

Self-Organizing Maps

Cao Mai

December 14, 2009

outline
Outline
  • Define Self-Organizing Maps (SOM)
  • Quick Example
  • Explain SOM
  • An example of SOM using MGP dataset
  • Questions
define
Define
  • A Self-Organizing Map (SOM) is a way to visually represent multi-dimensional data in an usually 2-D or 3-D manner
  • Similar data is grouped together.
  • Unsupervised algorithm
  • Continuously learning and adapting.
quick example
Quick Example
  • World Poverty Levels

Source: http://www.ai-junkie.com/ann/som/som5.html

randomized map generation
Randomized Map Generation
  • Randomized weights are assigned to each neuron

# of weights per neuron = # of attributes

Source: Pyle, D. (2003). Business Modeling and Data Mining. San Francisco: Morgan Kaufmann. Page 228

training the map
Training the Map
  • SOM instance recognition
  • Use Euclidean distance to match
    • This gives a measurement of how similar the two sets of data are to each other.

Source: Pyle, D. (2003). Business Modeling and Data Mining. San Francisco: Morgan Kaufmann. Page 228

altering neurons weights
Altering Neurons Weights
  • Both the Best Match Neuron and its surrounding neighbors weights are altered

Source: Pyle, D. (2003). Business Modeling and Data Mining. San Francisco: Morgan Kaufmann. Page 228

example with mpg dataset
Example with MPG dataset
  • MPG of 398 cars from 1970-1982
  • 6 Additional Attributes
    • Number of cylinders
    • Displacement
    • Horsepower
    • Weight
    • Acceleration
    • Model year

Except from the MPG dataset

results
Results

Engine Cylinder

results1
Results

Engine Cylinder

MPG

results2
Results

Engine Cylinder

Horsepower

slide12

Results

Engine Cylinder

Weight

slide13

Results

Engine Cylinder

Acceleration

slide14

Results

Engine Cylinder

Year

review
Review
  • Define SOM
  • Quick Example
  • Explain SOM
  • An example of SOM using MGP dataset
  • Questions
references
References
  • Aleshunas, John. Source Code & Executables. “Self-Organizing Map (SOM)” Retrieved: 2 Dec 2009 <http://mercury.webster.edu/Aleshunas/Canary/Canary_Home.html>
  • Pyle, D. (2003). Business Modeling and Data Mining. San Francisco: Morgan Kaufmann.
  • "Self-organizing map" Wikipedia, The Free Encyclopedia. Retrieved: 2 Dec. 2009, Wikimedia Foundation, Inc.. <http://en.wikipedia.org/wiki/Self-organizing_map>