No Access

A Self-Adaptive Weighted Fuzzy c-Means for Mixed-Type Data

School of Mathematics and Quantitative Economics, Shandong University of Finance and Economics, Jinan, Shandong Province, P. R. China

Corresponding author.

Search for more papers by this author

Zhihao Wang

School of Management and Engineering, Shandong University of Finance and Economics, Jinan, Shandong Province, P. R. China

Search for more papers by this author

, and

Guangfen Yang

School of Mathematics and Quantitative Economics, Shandong University of Finance and Economics, Jinan, Shandong Province, P. R. China

Search for more papers by this author

https://doi.org/10.1142/S1469026820500303Cited by:1 (Source: Crossref)

Abstract

The influence of features on each cluster is not the same in a mixed-type dataset. Based on the rough set and shadow set theories, the fuzzy distribution centroid was defined to represent the clustering center of the discrete feature so that the fuzzy c-means algorithm (FCM) could be extended to cluster the data with both continuous and discrete features. Then, considering the different contributions of the features to each cluster, a new weighted objective function was constructed in accordance with the principles of fuzzy compactness and separation. Because the learning feature weight is the key step in feature-weighted FCM, this paper regarded the feature weight as a variable optimized in the clustering process and put forward a self-adaptive mixed-type weighted FCM. The experimental results showed that the algorithm could be effectively applied to a heterogeneous mixed-type dataset.

Keywords:

Remember to check out the Most Cited Articles!
Check out these titles in artificial intelligence!

References

1. X. Wang et al., Feature selection based on rough sets and particle swarm optimization, Pattern Recogn. Lett. 28 (2007) 459–471. Crossref, Google Scholar
2. X. Wang, Y. Wang and L. Wang, Improving fuzzy c-means clustering based on feature-weight learning, Pattern Recogn. Lett. 25 (2004) 1123–1132. Crossref, Google Scholar
3. L. Wang, S. Guan, X. Wang and X. Wang, Fuzzy c mean algorithm based on feature weights, Chin. J. Comput. 29 (2006) 1797–1803. Google Scholar
4. Y. Wu, D. Hou, J. Li and S. Mi, Research on attribute-weighting cluster algorithm, J. Chin. Mini-Micro Comput. Syst. 33 (2012) 651–654. Google Scholar
5. J. Li, X. Gao and L. Jiao, A new feature weighted fuzzy clustering algorithm, Chin. J. Electron. 34 (2006) 89–92. Google Scholar
6. W. L. Hung, M. S. Yang and D. H. Chen, Bootstrapping approach to feature-weight selection in fuzzy c-means algorithms with an application in color image segmentation, Pattern Recogn. Lett. 29 (2008) 1317–1325. Crossref, Google Scholar
7. Q. Wang, Y. Ye and J. Z. Huang, Fuzzy k-means with variable weighting in high dimensional data analysis, in Proc. 9th Web-Age Information Management (Zhangjiajie China, 2008), pp. 365–372. Crossref, Google Scholar
8. M. Nazari, J. Shanbehzadeh and H. Sarrafzadeh, Fuzzy c-means based on automated variable feature weighting, Lect. Notes Eng. Comput. Sci. 2202 (2013) 25–29. Google Scholar
9. X. Kong and X. Tong, Adaptive algorithms for weight of the feature weighted of FCM, in Proc. 5th Int. Conf. on Image and Graphics (Xi’an, China, 2009), pp. 501–505. Crossref, Google Scholar
10. T. F. Ng, T. D. Pham and C. Sun, Automated feature weighting in fuzzy declustering-based vector quantization, in Proc. 20th Int. Conf. Pattern Recognition (Istanbul, Turkey, 2010), pp. 686–689. Crossref, Google Scholar
11. D. Li, H. Gu and L. Zhang, Fuzzy c-means algorithm with interval-supervised attribute weights, Control Decis. 25 (2010) 457–460. Google Scholar
12. H. Huang, C. Kong, H. Yu and F. Wen, Research on adaptive entropy weight fuzzy c-means clustering algorithm, Syst. Eng. Theor. Pract. 36 (2016) 219–223. Google Scholar
13. W. S. Desarbo et al., Synthesized clustering: A method for amalgamating alternative clustering bases with differential weighting of variables, Psychometrika 49 (1984) 57–78. Crossref, Google Scholar
14. X. Wang, Y. Wang and L. Wang, Improving fuzzy c-means clustering based on feature-weight learning, Pattern Recogn. Lett. 25 (2004) 1123–1132. Crossref, Google Scholar
15. H. J. Xing and M. H. Ha, Further improvements in feature-weighted fuzzy c-means, Inform. Sciences 267 (2014) 1–15. Crossref, Google Scholar
16. L. Q. Peng, On robust and high-dimensional data clustering, PhD thesis, Xidian University (2011). Google Scholar
17. J. Z. Huang et al., Automated variable weighting in k-means type clustering, IEEE Trans. Pattern Anal. 27 (2005) 657–668. Crossref, Google Scholar
18. P. Lingras and C. West, Interval set clustering of web users with rough k-means, J. Intell. Inf. Syst. 23 (2004) 5–16. Crossref, Google Scholar
19. S. Mitra, W. Pedrycz and B. Barman, Shadowed c-means: Integrating fuzzy and rough clustering, Pattern Recogn. 43 (2010) 1282–1291. Crossref, Google Scholar
20. J. Zhou, W. Pedrycz and D. Miao, Shadowed sets in the characterization of rough-fuzzy clustering, Pattern Recogn. 44(8) (2011) 1738–1749. Crossref, Google Scholar
21. J. C. Ji, Research on clustering algorithms for the data with multidimensional mixed attributes, PhD thesis, Jilin University (2011). Google Scholar
22. J. Ji et al., A fuzzy k-prototypes clustering algorithm for mixed numeric and categorical data, Neurocomputing 30 (2013) 129–135. Google Scholar
23. L. Wang, J. Wang, T. Li and F. Ye, Cluster’s feature weighting fuzzy clustering algorithm integrating rough sets and shadowed sets, Syst. Eng. Electron. 35 (2013) 1769–1776. Google Scholar
24. L. Wang, J. Wang and J. Jiang, New shadowed c-means clustering with feature weights, T. Nanjing U. Aeronaut. Astronaut. 29 (2012) 273–285. Google Scholar
25. Z. Huang, Clustering large data sets with mixed numeric and categorical values, in Proc. 1st Pacific-Asia Knowled. Discov. Data Mining (1997), 21–34. Google Scholar
26. Z. Huang, Extensions to the k-means algorithm for clustering large data sets with categorical values, Data Min. Knowl. Disc. 2 (1998) 283–304. Crossref, Google Scholar
27. J. Ji et al., An improved k-prototypes clustering algorithm for mixed numeric and categorical data, Neurocomputing 120 (2013) 590–596. Crossref, Google Scholar