SIMILARITY OF BI-VARIANT TIME SERIES

Abstract

This paper presents one algorithm for choosing optimal solution from a bi-variant time series database. The presented algorithm uses a similarity measure to compare each object in the database to a specific suitable object. The proposed method is useful when dealing with multidimensional time series data, e.g., bi-variant time series, where calculation of distances between all elements can be time-consuming for certain applications, e.g., in situations where real-time analysis of data is of critical importance.