Classification of multivariate time series using locality preserving projections
Introduction
Time series classification is an important problem in time series data mining. Many algorithms have been proposed for univariate time series classification [1], [2], [3], [4], [5], however, few papers are found about Multivariate time series (MTS) classification in the literature. MTS are used in very broad areas such as multimedia, medicine, finance and speech recognition. MTS classification is difficult for traditional machine learning algorithms mainly because of dozens of variables and different lengths of MTS samples.
MTS is a series of observations, xi(t)[i = 1, 2, … , n; t = 1, 2, … , m], where m is the number of observations and n is the number of variables (n is greater than, or equal to 2) [21]. MTS sample is stored by using m × n matrix. Classifying MTS samples poses two challenges [22]:
- –
MTS sample has dozens of variables. There are usually important correlations among the variables. If an MTS sample is broken into multiple univariate time series and each processed separately, the correlations among the variables in MTS sample will be lost.
- –
MTS samples can be of different lengths, even for similar MTS samples. For instance, the length of MTS sample in Japanese Vowels dataset [16] is between 7 and 29, the average length of MTS sample in Australian Sign Language dataset [16] is around 60.
Li et al. [6], [22] proposed two different feature vector selection approaches from MTS sample using the singular vector decomposition (SVD). The first approach takes into account the first singular vector and the normalized singular values, and the second approach considers the first two dominating singular vectors weighted by their associated singular values. The correlations among the variables can be extracted by Li’s two approaches. However, the class labels are not taken into consideration when feature vectors are extracted from MTS samples, and the dimensionality of these feature vectors is relatively high.
In this paper, a new approach to classifying MTS using locality preserving projections (LPP) [8], [9] is proposed. LPP is linear projective map that optimally preserves the neighborhood structure of the dataset. Our approach first extracts feature vectors from MTS samples using Li’s two different approaches [6], [22]; then by using LPP, these feature vectors are projected into a lower-dimensional space in which the MTS samples related to the same class are close to each other; finally, the MTS samples in testing set can be identified by one-nearest-neighbor (1NN) classifier in the lower-dimensional space. To the best of our knowledge, LPP has not been previously investigated for MTS classification.
In comparison with our proposed approach (Li’s two different feature vector selection approaches + LPP + 1NN), we also adopt two-dimensional singular value decomposition (2dSVD) [15] for feature extraction. A feature matrix is obtained for each MTS sample by using 2dSVD.
The contributions of this work include:
- –
Projecting feature vectors extracted from MTS samples into a lower-dimensional space by using LPP, in which the MTS samples related to the same class are close to each other.
- –
Applying one-nearest-neighbor classifier to the feature vectors of the MTS samples in the lower-dimensional space and classifying the MTS samples.
- –
Using 2dSVD to extract feature matrix from each MTS sample and comparing the 2dSVD + 1NN with our proposed approach.
Several experiments are performed on five real-world datasets. The results demonstrate the effectiveness of our approach in terms of accuracy and CPU time. Our approach significantly outperform Li’s two approaches + 1NN, outperform or perform comparably to 2dSVD + 1NN in terms of classification error rate and CPU time on five real-world datasets. The advantage of our approach is that our approach can be used with MTS samples of different lengths. However, 2dSVD assumes that the MTS samples have the equal length, which is very difficult to satisfy in real-world dataset.
The remainder of this paper is organized as follows. Section 2 gives a brief review of related work and background. Section 3 proposes a new approach for MTS classification using LPP. Section 4 experimentally demonstrates the effectiveness of our proposed approach. Conclusion is presented in Section 5.
Section snippets
Related works and background
This section gives a brief review of related works and some background knowledge of locality preserving projections (LPP) and two-dimensional singular value decomposition (2dSVD).
Classification of MTS using LPP
Classification of MTS Using LPP has three steps: in the first step, we calculate LPP from feature vectors extracted from training dataset; in the second step, by using LPP, feature vectors extracted from training dataset and testing dataset are projected into a lower-dimensional space; in the third step, the MTS samples in testing set can be identified by 1NN classifier with Euclidean distance in the lower-dimensional space. The algorithmic procedure is stated below:Algorithm 1. MTS_LPP_1NN
Experiments
In experiments, we adopt Li’s two approaches, Li’s two approaches + LPP, and 2dSVD for feature extraction, one-nearest-neighbor (1NN) classifier with Euclidean distance for classification, note that (5) is used to calculate the distance between two feature matrices extracted by using 2dSVD. The parameter t in Sij is set as 1.
In principle, any learner (SVM, kNN, Bayesian, etc.) could be used in this work. However, it has been shown that 1NN with Euclidean distance is very hard to beat [23], [24].
Conclusion
In this paper, we have presented a new approach to handle MTS classification problem. Our approach is based on Li’s two different feature vector selection approaches and LPP. By using LPP, feature vectors extracted from MTS samples are projected into a lower-dimensional space in which the MTS samples related to the same class are close to each other. The MTS samples in testing set can be identified by a one-nearest-neighbor classifier in the lower-dimensional space. By choosing the appropriate
Acknowledgments
We thank the maintainers of the UCI KDD Repository [16] and the donors of the different datasets [18], [26]. This work was supported by the National Science Foundation of China under the Grant No. 60173058.
References (26)
- et al.
Image retrieval based on incremental subspace learning
Pattern Recognition
(2005) Comparison and classification of stationary multivariate time series
Pattern Recognition
(1999)- L. Chen, M. Kamel, J. Jiang, A modular system for the classification of time series data, in: Proc. MCS 2004, LNCS...
- H. Zhang, T. Ho, W. Huang, Blind feature extraction for time series classification using Haar wavelet transform, in:...
- et al.
Support vector machines of interval-based features for time series classification
Knowledge-Based Systems
(2005) - et al.
Time series classification using gaussian mixture models of reconstructed phase spaces
IEEE Transactions on Knowledge and Data Engineering
(2004) - A. Hayashi, Y. Mizuhara, N. Suematsu, Embedding time series data for classification, in: Proc. MLDM 2005, LNAI 3587,...
- et al.
Real-time classification of variable length multi-attribute motion data
International Journal of Knowledge and Information Systems
(2006) Classification of multivariate time series and structured data using constructive induction
Machine Learning
(2005)- et al.
Locality preserving projections
(2003)