MULTI-VARIATE TIME SERIES CLUSTERING BASED ON TWO-DIMENSIONAL SINGULAR VALUE DECOMPOSITION AND GAUSSIAN MIXTURE MODEL
Multi-variable time series(MTS)has two dimensions,time and variable,and traditional principal component analysis(PCA)is limited in MTS data representation.Therefore,a MTS clustering algorithm based on the two-dimensional singular value decomposition(2dSVD)and Gaussian mixture model(GMM)is proposed.We calculated the eigenvectors of the row-row and column-column covariance matrix of MTS,extracted the eigenmatrix from the two dimensions of time and variable,and then used GMM to cluster the eigenmatrix from the perspective of probability distribution.The experimental results demonstrate that the new algorithm can gain better clustering results on MTS.
Two-dimensional singular value decompositionGaussian mixture modelMulti-variate time series clustering
万方数据
维普
