Direkt zum Inhalt
 
 
Eine algebraische Fläche vom Grad 6 (eine "Sextik"), die 65 Singularitäten besitzt.
 
  Startseite  
 

Fundamental Clustering Problem Suite

The Fundamental Clustering Problems Suite (FCPS) offers a variety of clustering problems any algorithm shall be able to handle when facing real world data. FCPS serves as an elementary benchmark for clustering algorithms.

FCPS consists of data sets with known a priori classifications that are to be reproduced by the algorithm. All data sets are intentionally created to be simple and might be visualized in two or three dimensions. Each data sets represents a certain problem that is solved by known clustering algorithms with varying success. This is done in order to reveal benefits and shortcomings of algorithms in question. Standard clustering methods, e.g. single-linkage, ward und k-means, are not able to solve all FCPS problems satisfactorily.

FCPS is supposed to be used in scientific works for free, as long as it is quotet as follows:

Ultsch, A.: Clustering with SOM: U*C, In Proc. Workshop on Self-Organizing Maps, Paris, France, (2005) , pp. 75-82

click here for data (.zip file)

Name

Problem

Description

Image

U-Matrix

P-Matrix 

Hepta clearly defined clusters, different variances

Size 212

Dimensions 3

Classes 7 

 tinyHepta   tinyHeptaUmx  tinyHeptaPmx
Lsun different variances and inter cluster distances

Size 400

Dimensions 2

Classes 3

 tinyLsun  tinyLsunUmx

 tinyLsunPmx

 Tetra almost touching clusters

Size 400

Dimensions 3

Classes 4

 tinyTetra  tinyTetraUmx  tinyTetraPmx
Chainlink linear not separable

Size 1000

Dimensions 3

Classes 2

 tinyChainlink  tinyChainlinkUmx  tinyChainlinkPmx
Atom different variances and linear not separable

Size 800

Dimensions 3

Classes 2

 tinyAtom  tinyAtomUmx  tinyAtomPmx
EngyTime Gaussian mixture

Size 4096

Dimensions 2

Classes 2

 tinyEngyTime  tinyEngyTimeUmx  tinyEngyTimePmx
Target outliers

Size 770

Dimensions 2

Classes 6

 tinyTarget  tinyTargetUmx  tinyTargetPmx
TwoDiamonds cluster borders defined by density

Size 800

Dimensions 2

Classes 2

 tinyTwoDiamonds  tinyTwoDiamondsUmx  tinyTwoDiamondsPmx
Wingnut density vs. distance

Size 1070

Dimensions 2

Classes 2

 tinyWingnut  tinyWingnutUmx  tinyWingnutPmx
Golfball no clusters at all

Size 4002

Dimensions 3

Classes 2

 tinyGolfball  tinyGolfballUmx  tinyGolfballPmx


Zuletzt aktualisiert: 16.12.2013 · herrmanl

 
 
 
Fb. 12 - Mathematik und Informatik

Datenbionik (AG Ultsch), Hans-Meerwein-Straße, D-35032 Marburg
Tel. +49 6421/28-22185, Fax +49 6421/28-28902, E-Mail: databionics@informatik.uni-marburg.de

URL dieser Seite: http://www.uni-marburg.de/fb12/datenbionik/data

Impressum | Datenschutz