UH-DMML Group

2D Spatial Datasets

Author: Sujing Wang, Nidal Zeidat, Christoph F. Eick

Note:    The Complex9, Complex8 and Diamond9 datasets were obtained from:

Salvador, S. and Chan, P., ”Determining the Number of Clusters/Segments in Hierarchical clustering/Segmentation Algorithm”, ICTAI 2004,576-584.

 

We also created 6 more datasets by adding 8%, 16%, and 32%, random and Gaussian noise to the Complex9 dataset.

Visualization of 2D Spatial Datasets

1.      Complex9 Based Datasets:

 

Dataset information:

·        Number of Instances: 3031

·        Number of Attributes: 2 numeric (X,Y coordinate) and the class

·        class: total 9 classes(from 0 to 8)

·        Missing Attribute Values: None

·        Format:  x, y, class-label

·        download file: complex9.data

 

 

2.      Complex8 dataset:

 

Dataset information:

• Number of Instances: 2551
• Number of Attributes: 2 numeric (X,Y coordinate) and the class
• class: total 8 classes(from 0 to 7)
• Missing Attribute Values: None
• Format:  x, y, class-label
• download file: complex8.data

 

3.      Diamond9 dataset:

 

Dataset information:

• Number of Instances: 3000
• Number of Attributes: 2 numeric (X,Y coordinate) and the class
• class: total 9 classes(from 0 to 8)
• Missing Attribute Values: None
• Format:  x, y, class-label
• download file: diamond9. data