25th annual Conference of the German Classification Society(GfKl)
University of Munich 14-16 March 2001

Mining Sets of time Series: description of time points

Aldar Dugarjapov, Georg Lausen

Abstract:

Investigating time series with respect to statistical analysis and forecasting is a well established area. In this paper we study the problem of matching time series. This paper addresses the problem how to choose the attributes describing the time points with respect to certain matching problem. Furthemore, this paper introduce an algorithm, computing the domain error of a classifier of the model, which can be used as an evaluation function of the classifier with respect to the matching problem. The exeprimental results, presented in this paper, illustrate the usefulness of the evaluation.

[PDF File] [PS File]