USING OF CLUSTERING METHODS IN INFOGRAPHIC PROBLEMS

The article is devoted to using of various generalizations of the k-means method in infographic problems. It is shown that for effective visualization of each task it is necessary to use the quality criterion most suitable for displaying the specificity of the data used. For a number of problems, it is suggested to use the gravitational model instead of the Euclidean distance or Mahalonobis distance. It is proposed to use the shuffled frog-leaping algorithm to optimize the operation of the algorithm.