Online Recognition of Fuzzy Time Series Patterns

Abstract

This article deals with the recognition of recurring\nmultivariate time series patterns modelled sample-point-wise by parametric\nfuzzy sets. An efficient classification-based approach for the\nonline recognition of incompleted developing patterns in streaming\ntime series is being presented. Furthermore, means are introduced\nto enable users of the recognition system to restrict results to certain\nstages of a pattern’s development, e. g. for forecasting purposes, all\nin a consistently fuzzy manner.