Algorithms for time series comparison
Jiří Fejfar, Arnošt Motyčka, Štěpán Filípek · 2011
In this paper we compare results of four chosen algorithms suitable for the time series clustering. The comparison is made in sense of agreement between clustering results represented in a special form of confusion matrix (matching matrix). We use a musical data, specifically music excerpts volume development, as a time series instances. This musical characteristic is used in computer-aided tool for musical analysis developed by our group.