On reconstruction of chaotic attractor from time series represented as "clusters"
Abstract
The method of reconstruction of an attractor from a set of short time series ({\it clusters}) is proposed and discussed. This method is most useful for correlation dimension estimation of experimental data.