MAS 2012 Proceeding

Unsupervised algorithm for retrieving characteristic patterns from time-warped data collections

Authors:   Tomá? Kocyan, Jan Martinovi?, Michal Podhorányi, Ivo Vondrák


This paper discusses possibilities of using the Voting Experts algorithm enhanced by the Dynamic Time Warping (DTW) method for improving performance of Case-Based Reasoning (CBR) methodology used with time-warped data collections. CBR, in general, is the process of solving new problems based on the solutions of similar past problems. Success of this methodology strongly depends on the ability to find similar past situations. Searching these similar situations in data collections with components generated in equidistant time and in finite number of levels is now a trivial task. The problem arises for data collections that are subject to different types of distortions (e.g. measurement of natural phenomena such as precipitations, measured discharge volume etc.). The main goal of this paper is to provide suitable mechanism for retrieving typical patterns from distorted time series and thus improve the usability of CBR.

I3M  Scientific Sponsors

I3M  Industrial Sponsors

I3M  Media Sponsors