Implementing Correlation Dimension: K-Means Clustering via Correlation Dimension
The estimation of intrinsic dimension is an essential step in the dimension reduction process. The intrinsic dimension can be estimated by fractal dimension estimation methods, which exploit the intrinsic geometry of a data set. The most popular concept from this family of methods is the correlation dimension. K-Means is most popular clustering algorithm for performing unsupervised learning in data mining. This paper we propose approaches to approximate the correlation integral. In addition, we propose a new selection for clusters via correlation dimension. The performance of the algorithm is discussed. Experimental results on an artificial and real-world data are used to demonstrate the algorithms and compare to other methodology.
KeywordsIntrinsic dimension Fractal dimension Correlation dimension Clustering K-Means clustering
- 1.Camastra, F., Vinciarelli, A.: Estimating the intrinsic dimension of data with a fractal-based method. IEEE Trans. Pattern Anal. Mach. Intell. 24(10), 1404–1407 (2002)Google Scholar
- 6.Rencher, A., Christensen, W.: Methods of Multivariate Analysis. 3rd edn. Willey (2012)Google Scholar
- 7.Kalantan, Z., Einbeck, J.: On the computation of the correlation integral for fractal dimension estimation. In: International Conference on Statistics in Science. Business and Engineering (ICSSBE 2012), IEEE Conference Publications, pp. 80–85(2012). https://doi.org/10.1109/icssbe.2012.6396531