Multivariate time series classification: A relational way

Abstract

Multivariate Time Series Classification (MTSC) has attracted increasing research attention in the past years due to the wide range applications in e.g., action/activity recognition, EEG/ECG classification, etc. In this paper, we open a novel path to tackle with MTSC: a relational way. The multiple dimensions of MTS are represented in a relational data scheme, then a propositionalisation technique (based on classical aggregation/selection functions from the relational data field) is applied to build interpretable features from secondary tables to “flatten” the data. Finally, the MTS flattened data are classified using a selective Naïve Bayes classifier. Experimental validation on various benchmark data sets show the relevance of the suggested approach.