Modified A-LTK : Improvement of a Multi-dimensional Time Series Classification Method

— In a previous study, we proposed a novel method for approximating multi-dimensional time-series, called multi-dimensional time-series Approximation with use of Local features at Thinned-out Keypoints (A-LTK), which was shown to obtain a sufficiently accurate value when reduced storage cost is a requirement. In this study, we propose a modified version of this method where two changes are made to address the problem of degraded accuracy caused by high dimensionality. Our evaluation indicates that the Modified A-LTK is capable of achieving similar or superior accuracy compared with A-LTK and other existing methods but with the added advantage of reduced processing costs.

[1]  Dimitrios Gunopulos,et al.  Discovering similar multidimensional trajectories , 2002, Proceedings 18th International Conference on Data Engineering.

[2]  W. Timothy Liu,et al.  Tropical cyclone event sequence similarity search via dimensionality reduction and metric learning , 2010, KDD '10.

[3]  Dimitrios Gunopulos,et al.  Indexing Multidimensional Time-Series , 2004, The VLDB Journal.

[4]  Eamonn J. Keogh,et al.  Locally adaptive dimensionality reduction for indexing large time series databases , 2001, SIGMOD '01.

[5]  Kuniaki Uehara,et al.  Discovery of Time-Series Motif from Multi-Dimensional Data Based on MDL Principle , 2005, Machine Learning.

[6]  Anju Vyas.  Print , 2003 .

[7]  Majid Sarrafzadeh,et al.  Toward Unsupervised Activity Discovery Using Multi-Dimensional Motif Detection in Time Series , 2009, IJCAI.

[8]  Paul J. M. Havinga,et al.  Activity Recognition Using Inertial Sensing for Healthcare, Wellbeing and Sports Applications: A Survey , 2010, ARCS Workshops.

[9]  Oscar Castillo,et al.  Proceedings of the International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007, March 21-23, 2007, Hong Kong, China , 2007, IMECS.

[10]  Yixin Chen,et al.  Multi-Dimensional Regression Analysis of Time-Series Data Streams , 2002, VLDB.

[11]  Dimitrios Gunopulos,et al.  Indexing multi-dimensional time-series with support for multiple distance measures , 2003, KDD '03.

[12]  Eamonn J. Keogh,et al.  Dimensionality Reduction for Fast Similarity Search in Large Time Series Databases , 2001, Knowledge and Information Systems.

[13]  Hung-Hsuan Huang,et al.  Multi-dimensional Time Series Approximation Using Local Features at Thinned-out Keypoints , 2015, J. Comput..

[14]  Lei Chen,et al.  On The Marriage of Lp-norms and Edit Distance , 2004, VLDB.

[15]  M. Reinders,et al.  Multi-Dimensional Dynamic Time Warping for Gesture Recognition , 2007 .

[16]  Anthony J. T. Lee,et al.  Mining closed patterns in multi-sequence time-series databases , 2009, Data Knowl. Eng..