pyclust-evl 0.1.0

A Python library for clustering operations. Evaluation and meta-feature generation.

The PyClustKit Module: All about clustering in a single Python Module!

The pyclustkit module is built on top of various libraries to enable many clustering operations. Currently, the module is built for clustering evaluation and meta-learning.

Table of Contents

• Installation Instructions
• Useful Links
• Usage Examples
• Cite Us!

Installation Instructions

The pyclustkit is available to download with pypi

pip install pyclustkit

I

Useful Links

Usage Examples

Calculating Internal Cluster Validity Indices (CVI)

PyClustKit comes with an evaluation suite of 46 internal validity indices. Each is implemented on top of numpy and, the module incorporates specific methods for speeding up the execution of multiple CVI by implementing a shared process tracking.

from pyclustkit.eval import CVIToolbox 

ct = CVIToolbox(X,y)
ct.calculate_icvi(cvi=["dunn", "silhouette"]) # if no CVI are specified it defaults to 'all'.
print(ct.cvi_results)

Meta Learning

Meta-Feature Extraction

PyClustKit comes with an evaluation suite of 46 internal validity indices. Each is implemented on top of numpy and, the module incorporates specific methods for speeding up the execution of multiple CVI by implementing a shared process tracking.

from pyclustkit.eval import CVIToolbox 

ct = CVIToolbox(X,y)
ct.calculate_icvi(cvi=["dunn", "silhouette"]) # if no CVI are specified it defaults to 'all'.
print(ct.cvi_results)

Citing This Work

List of Implemented CVI with citations

Currently the collection consists of the following internal CVIs. R does not do gdi 61,62,63 due to hausdorff:

1. ball_hall: G. H. Ball and D. J. Hall. Isodata: A novel method of data analysis and pattern classification. Menlo Park: Stanford Research Institute. (NTIS No. AD 699616),1965.

2. banfeld_raftery: J.D. Banfield and A.E. Raftery. Model-based gaussian and non-gaussian clustering. Biometrics, 49:803–821, 1993.

3. c_index: Hubert, Lawrence & Levin, Joel. (1976). A general statistical framework for assessing categorical clustering in free recall. Psychological Bulletin. 83. 1072-1080. 10.1037/0033-2909.83.6.1072.

4. CDbw : Halkidi, M., & Vazirgiannis, M. (2008). A density-based cluster validity approach using multi-representatives. Pattern Recognit. Lett., 29, 773-786.

5. det_ratio : A. J. Scott and M. J. Symons. Clustering methods based on likelihood ratio criteria. Biometrics, 27:387–397, 1971.

6. Dunn Index : J. Dunn. Well separated clusters and optimal fuzzy partitions. Journal of Cybernetics, 4:95–104, 1974.

7. GDI [11,21,31,41,51,61][12,22,32,42,52,62][13,23,33,43,53,63]: J. C. Bezdek and N. R. Pal. Some new indexes of cluster validity. IEEE Transactions on Systems, Man, and CyberneticsÑPART B: CYBERNETICS, 28, no.3:301–315, 1998.

8. ksq_detw: F. H. B. Marriot. Practical problems in a method of cluster analysis. Biometrics, 27:456–460, 1975.

9. log_det_ratio: Halkidi et al. On clustering validation techniques. J. Intell. Inf. Syst., 2001.

10. log_ss_ratio: J. A. Hartigan. Clustering algorithms. New York: Wiley, 1975.

11. McClain_Rao: J. O. McClain and V. R. Rao. Clustisz: A program to test for the quality of clustering of a set of objects. Journal of Marketing Research, 12:456–460, 1975.

12. trace_w Index

13. Friedman-Rudin 1 Index

14. Friedman-Rudin 2 Index

15. S_dbw: M. Halkidi and M. Vazirgiannis, "Clustering validity assessment: finding the optimal partitioning of a data set," Proceedings 2001 IEEE International Conference on Data Mining.

16. sd_dis Index: Halkidi et al. On clustering validation techniques. J. Intell. Inf. Syst., 2001.

17. sd_scat Index: Halkidi et al. On clustering validation techniques. J. Intell. Inf. Syst., 2001.

18. pbm: Bandyopadhyay S. Pakhira M. K. and Maulik U. Validity index for crisp and fuzzy clusters. Pattern Recognition, 2004.

19. ratkowsky_lance

21. ray_turi: Ray et al. Determination of number of clusters in k-means clustering and application in colour image segmentation. 4th International Conference on Advances in Pattern Recognition and Digital Techniques, 1999.

22. wemmert_gancarski

23. xie_beni: X.L. Xie and G. Beni. A validity measure for fuzzy clustering. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991.

25. banfeld_raftery

26. trace_wib

28. log_det_ratio

30. point_biserial

31. calinski_harabasz

32. silhouette

33. davies_bouldin

34. scott_symons