We present a new approach to the evaluation of similarity of time series that are characterized by linguistic summaries. We consider so-called temporal data summaries, i.e. novel linguistic summaries that explicitly include a temporal aspect. We consider the case of a mutual (investment) fund and its underlying benchmark(s), and the new comparison method is based not on the comparison of the consecutive values or segments of the fund and its benchmark but on the comparison of classic and temporal linguistic summaries (i.e. based on a classic and temporal protoform) best describing their past behavior.
|Title of host publication||Combining soft computing and statistical methods in data analysis|
|Editors||C. Borgelt, G. Gonzalez-Rodriguez, W. Trutschnig, M.A. Lubiano, M.A. Gil, P. Grzegorzewski, O. Hryniewicz|
|Place of Publication||Berlin|
|Number of pages||9|
|Publication status||Published - 2010|
|Name||Advances in Intelligent and Soft Computing (AINSC)|