SARC INTERNATIONAL CONFERENCE – Proceedings
Papers
-
Classification of Time Series through Ensembles of Different Distance Measures
Tri Pham, Anh Duong Tuan
Paper ID SA-MLAI-VTNM-070521-2412 · Pages 32–63
Download PDFAbstract
This paper presents an ensemble approach for time series classification which is based on elastic and non-elastic distance measures for a 1-NN (one nearest neighbor) method. We also design a multi-threaded implementation for the proposed ensemble method in order to improve its efficiency. Experimental results over a large collection of benchmark datasets showed that our proposed method remarkably outperforms the 1NN with Dynamic Time Warping measure which has been considered in literature as “difficult to beatâ€. Besides, we compare the efficiency of multi-threaded implementation to that of sequential implementation for the proposed method. Results in the latter experiment showed that in average the multi-threaded version can run faster that the sequential version about 2.8 times. Keywords - Time Series, Classification, Ensemble, Elastic Distance Measure, Non-elastic Distance Measure.