出版社：Vilnius University, University of Latvia, Latvia University of Agriculture, Institute of Mathematics and Informatics of University of Latvia
摘要：Multidimensional data are often difficult to understand for a human because of their high dimensionality. Multidimensional data visualization is one of the ways for data perception where multidimensional data must be transformed in a low-dimensional space and presented visually for human decision. As a result of transformation there appear new data features, the number of which is lower than that of the original data features. In this paper, we present and investigate the way of reduction of dimensionality using the exponential correlation function, taking into account that there are clusters in the analysed set of multidimensional data.
