Abstract
Echo state network (ESN) is a novel kind of recurrent neural networks, where a reservoir is generated randomly and only readout layer is adaptable. It outperforms conventional recurrent neural networks in the field of multivariate time series prediction. Often ESN works beautifully. But sometimes it works poorly because of ill-posed problem. To solve it, we propose a new model on the basis of ESNs, termed as fast subspace decomposition echo state network (FSDESN). The core of the model is to utilize fast subspace decomposition algorithm for extracting a compact subspace out of a redundant large-scale reservoir matrix in order to remove approximate collinear components, overcome the ill-posed problem, and improve generalization performance. Experimental results on two multivariate benchmark datasets substantiate the effectiveness and characteristics of FSDESN.
Similar content being viewed by others
References
Cordova J, Yu W (2012) Two types of Haar wavelet neural networks for nonlinear system identification. Neural Process Lett 35:283–300
Sapankevych N, Sankar R (2009) Time series prediction using support vector machines: a survey. IEEE Comput Intell Mag. doi:10.1109/MCI.2009.932254
Chen SM, Chang YC (2011) Weighted fuzzy rule interpolation based on GA-based weight-learning techniques. IEEE Trans Fuzzy Syst 19(4):729–744
Gil PG, Cortes JMR, Hernández SEP, Aquino VA (2011) A neural network scheme for long-term forecasting of chaotic time series. Neural Process Lett 33:215–233
Li TS, Feng G, Wang D, Tong SC (2010) Neural-network-based simple adaptive control of uncertain multi-input multi-output non-linear systems. IET Control Theory Appl 4(5):1543–1557
Li TS, Wang D, Feng G, Tong SC (2010) A DSC approach to robust adaptive NN tracking control for strict-feedback nonlinear systems. IEEE Trans Syst Man Cybern B Cybern 40(3):915–927
Wen SP, Zeng ZG, Huang TW (2012) Exponential stability analysis of memristor-based recurrent neural networks with time-varying delays. Neurocomputing 97:233–240
Jaeger H, Haas H (2004) Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 304:78–80
LukošEvičIus M, Jaeger H (2009) Reservoir computing approaches to recurrent neural network training. Comput Sci Rev 3:127–149
Rodan A, Tiňo P (2011) Minimum complexity echo state network. IEEE Trans Neural Netw 22(1):131–144
Shi ZW, Han M (2007) Support vector echo-state machine for chaotic time-series prediction. IEEE Trans Neural Netw 18(2):359–372
Chatzis SP, Demiris Y (2011) Echo state Gaussian process. IEEE Trans Neural Netw 22(9):1435–1445
Leung CS, Lam PM (2003) A local training–pruning approach for recurrent neural networks. Int J Neural Syst 13:25–38
Friedman JH (2012) Fast sparse regression and classification. Int J Forecast 28:722–738
Dutoit X, Schrauwen B, Campenhout JV, Stroobandt D, Brussel HV, Nuttin M (2009) Pruning and regularization in reservoir computing. Neurocomputing 72:1534–1546
Reichel L, Sgallari F, Ye Q (2012) Tikhonov regularization based on generalized Krylov subspace methods. Appl Numer Math 62(9):1215–1228
Tang X, Han M (2009) Partial Lanczos extreme learning machine for single-output regression problems. Neurocomputing 72:3066–3076
Mahadevan S (2008) Fast spectral learning using Lanczos eigenspace projections. In: Proceedings of the twenty-third AAAI conference on artifical intelligence, Chicago, pp 1472–1474
Xu G, Kailath T (1994) Fast subspace decomposition. IEEE Trans Signal Process 42(3):539–551
Rubio F, Mestre X (2009) Generalized consistent estimation on low-rank Krylov subspaces of arbitrarily high dimension. IEEE Trans Signal Process 57(10):3787–3800
Han M, Wang Y (2009) Prediction of multivariate time series based on reservoir principal component analysis. Control Decis 24(10):1526–1530
Acknowledgments
This research is supported by the project (61074096) of the National Natural Science Foundation of China.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Han, M., Xu, M. Predicting Multivariate Time Series Using Subspace Echo State Network. Neural Process Lett 41, 201–209 (2015). https://doi.org/10.1007/s11063-013-9324-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11063-013-9324-7