LDMLT_Multivariate_​Time_Series_Classif​ication.zip

This is the code of a novel metric learning algorithm for Multivariate Time Series Classfication.

Multivariate time series (MTS) data sets broadly exist in numerous fields, including health care, multimedia, finance and biometrics. How to classify MTS accurately has become a hot research point since it is an important element in many computer vision and pattern recognition applications. In the code, we propose a Mahalanobis distance based Dynamic Time Warping (MDDTW) measure for MTS classification. The Mahalanobis distance builds an accurate relationship between each variable and its corresponding category. It is utilized to calculate the local distance between vectors in MTS. Then we use Dynamic Time Warping (DTW) to align those MTS which are out of sync or with different lengths. Meanwhile, we use a LogDet divergence based metric learning with triplets constraints (LDMLT) model to the learn Mahalanobis matrix with high precision and robustness. Furthermore, we demostrate the perforamce of the code on MTS data "JapaneseVowels".