Fuzzy spacial extrapolation method using Manhattan metrics for tasks of Medical Data mining

In this paper the approach for fuzzy clustering-classification of medical short data samples using the method of fuzzy spatial extrapolation is considered. The proposed procedure refers to the direction of Medical Data Mining, and is hybrid system that can solve the task of diagnosing of various diseases in a limited sample, complete or partial overlapping of classes, their different densities, different numerical filling and requires for its training small volumes of a priori information. Also this procedure can realize a filling of gaps in feature vector based on recovery of hidden dependencies that are contained in data set.