The description problem of the clusters´ centroids

The paper is devoted to analysis of the forecasting method for time series groups with the use of k-means clustering algorithm. The coordinates of the centers of clusters have been put in correspondence with summarizing time series data - the centroids of the clusters. A description of time series, the centroids of the clusters, is implemented with the use of forecasting models. It is suggested to use a common forecasting model, which is constructed for the cluster´s centroid, in forecasting the individual time series in the cluster. The comparative analysis of the forecasting results with use of the classical Euclidean metric and the offered modified Euclid´s metrics in the k-means clustering algorithm is made. The advantage of the offered modified Euclidean metric the metrics is shown.