aydos.com

Fuzzy Clustering and Data Analysis

Find your system's clusters with two different fuzzy clustering algoritms (Fuzzy C-Means; Gustafson Kessel) and analyse with Least Square Estimation or Fuzzy Functions. You can analyse up to 8 (7 inputs, 1 output) dimensions.

How to use

• Load your data using Load Data window. After paste data into the text area, click Accept button. Application will review and accept max 2000 lines and 8 columns of your data. You can review what application has as data.
• Choose have may rows of data will be use as validation data. Train data is must. So you can choose max 40 lines of validation data.
• Choose how many center are in the data.
• Choose fuzziness coefficient. Best values are between 1.4 and 2.8 .
• Choose method to analyse.
• Click Calculate button.
• Get the results by Results button.

Links

• Fuzzy C-Means tutorial

Version History

2.4 (2014-01-05)

• Ported to aydos.com

2.3 (2011-10-09)

• Calculation for one cluster added

2.2 (2011-08-01)

• Small changes

2.1 (2011-02-21)

• Min-max of graph
• Clustering method: Gustafson-Kessel

2.0 (2011-02-16)

• User friendly graphic interface, tabs
• Graphs of data, clusters, memberships and estimates
• Visitor can analyse data of size 4x400; members can 6x1000
• Comparation current calculation with previous one

1.0 (2010-08-02)

• Basic graphic interface
• Clustering method: FCM (Fuzzy c-mean clustering)
• Analysis method: Fuzzy Least Square Estimation
• Analysis method: Fuzzy Functions

Acknowledgements

• Prof.Dr. Lotfi A. Zadeh
• Prof.Dr. Burhan Türkşen
• Minimal Comps by Keith Peters