Abstract: The clustering problem for multivariate observations often encountered in many applications connected with Data Mining and Exploratory Data Analysis. Conventional approach to solving these problems requires that each observation may belong to only one cluster, although a more natural situation is when the vector of features with different levels of probabilities or possibilities can belong to several classes. This situation is subject of consideration of fuzzy cluster analysis, intensively developing today. In many practical tasks of Data Mining, including clustering, data sets may contain gaps, information in which, for whatever reasons, is missing. More effective in this situation are approaches based on the mathematical apparatus of Computational Intelligence and first of all artificial neural networks and different modifications of classical fuzzy c-means (FCM) method. But these methods are effective only in cases when the original data set is given beforehand and does not change during data processing. At the same time there is a wide class of problems when the data are fed to processing sequentially in on-line mode as it occurs in self-organizing Kohonen networks training. At the same time apriori it is not known which of the vectors-images contain gaps. In this paper the problem of probabilistic and possibilistic on-line clustering of data with gaps using Partial Distance Strategy is discussed and solved, self-organizing neuro-fuzzy Kohonen network and new self-learning algorithm that is the hybrid of "Winner-takes-more" rule and recurrent fuzzy clustering procedures are proposed and investigated.

Keywords: Fuzzy clustering, Kohonen self-organizing network, learning rule, incomplete data with gaps.

ACM Classification Keywords: 1.2.6 Artificial Intelligence: Learning – Connectionism and neural nets; 1.2.8 Artificial Intelligence: Problem Solving, Control Methods, and Search – Control theory; 1.5.1 Pattern Recognition: Clustering – Algorithms.

Link:

ADAPTIVE CLUSTERING OF INCOMPLETE DATA USING NEURO-FUZZY KOHONEN NETWORK

Yevgeniy Bodyanskiy, Alina Shafronenko, Valentyna Volkova

http://www.foibg.com/ibs_isc/ibs-26/ibs-26-p20.pdf