A Neural Network Scheme for Long-Term Forecasting of Chaotic Time Series

The accuracy of a model to forecast a time series diminishes as the prediction horizon increases, in particular when the prediction is carried out recursively. Such decay is faster when the model is built using data generated by highly dynamic or chaotic systems. This paper presents a topology and training scheme for a novel artificial neural network, named “Hybrid-connected Complex Neural Network” (HCNN), which is able to capture the dynamics embedded in chaotic time series and to predict long horizons of such series. HCNN is composed of small recurrent neural networks, inserted in a structure made of feed-forward and recurrent connections and trained in several stages using the algorithm back-propagation through time (BPTT). In experiments using a Mackey-Glass time series and an electrocardiogram (ECG) as training signals, HCNN was able to output stable chaotic signals, oscillating for periods as long as four times the size of the training signals. The largest local Lyapunov Exponent (LE) of predicted signals was positive (an evidence of chaos), and similar to the LE calculated over the training signals. The magnitudes of peaks in the ECG signal were not accurately predicted, but the predicted signal was similar to the ECG in the rest of its structure.

[1]  M. Rosenstein,et al.  A practical method for calculating largest Lyapunov exponents from small data sets , 1993 .

[2]  Barak A. Pearlmutter.  Dynamic recurrent neural networks , 1990 .

[3]  Antonette Marie Logar.  Recurrent neural networks and time series prediction , 1992 .

[4]  Gustavo Deco,et al.  Dynamic modeling of chaotic time series by neural networks , 1997, Annual Conference Computational Learning Theory.

[5]  S. Duhau,et al.  An Early Prediction of Maximum Sunspot Number in Solar Cycle 24 , 2003 .

[6]  Jose C. Principe,et al.  Neural and Adaptive Systems: Fundamentals through Simulations with CD-ROM , 1999 .

[7]  José Carlos Príncipe,et al.  Dynamic Modelling of Chaotic Time Series with Neural Networks , 1994, NIPS.

[8]  Pilar Gómez-Gil,et al.  Time series forecasting using recurrent neural networks and wavelet reconstructed signals , 2010, 2010 20th International Conference on Electronics Communications and Computers (CONIELECOMP).

[9]  W. Kinsner.  Characterizing chaos through Lyapunov metrics , 2006, IEEE Trans. Syst. Man Cybern. Syst..

[10]  Joshua R. Smith,et al.  Preventing tomorrow's sudden cardiac death today: part I: Current data on risk stratification for sudden cardiac death. , 2007, American heart journal.

[11]  Robert Kozma,et al.  Time series prediction using chaotic neural networks on the CATS benchmark , 2007, Neurocomputing.

[12]  Inés María Galván,et al.  Multi-step Learning Rule for Recurrent Neural Models: An Application to Time Series Forecasting , 2004, Neural Processing Letters.

[13]  Niels Wessel,et al.  Ventricular arrhythmias and changes in heart rate preceding ventricular tachycardia in patients with an implantable cardioverter defibrillator , 2008, Medical & Biological Engineering & Computing.

[14]  V. Alarcon-Aquino,et al.  Multiresolution FIR neural-network-based learning algorithm applied to network traffic prediction , 2006, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[15]  Jeffrey L. Elman,et al.  Finding Structure in Time , 1990, Cogn. Sci..

[16]  Johan A. K. Suykens,et al.  Time Series Prediction Competition , 1999 .

[17]  Danilo P. Mandic,et al.  Recurrent Neural Networks for Prediction: Learning Algorithms, Architectures and Stability , 2001 .

[18]  Michael I. Jordan,et al.  Advances in Neural Information Processing Systems 30 , 1995 .

[19]  I. J. Leontaritis,et al.  Input-output parametric models for non-linear systems Part II: stochastic non-linear systems , 1985 .

[20]  Pilar Gómez-Gil,et al.  On the use of hybrid neural networks and non-linear invariants for prediction of electrocardiograms , 1999, IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339).

[21]  R. Gencay,et al.  An algorithm for the n Lyapunov exponents of an n -dimensional unknown dynamical system , 1992 .

[22]  M. Small,et al.  Towards long-term prediction , 2000 .

[23]  D. Wunsch,et al.  Time series prediction with recurrent neural networks using a hybrid PSO-EA algorithm , 2004, 2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541).

[24]  Mattias Ohlsson,et al.  In search of the best method to predict acute coronary syndrome using only the electrocardiogram from the emergency department. , 2009, Journal of electrocardiology.

[25]  Ravi Sankar,et al.  Time Series Prediction Using Support Vector Machines: A Survey , 2009, IEEE Computational Intelligence Magazine.

[26]  Jose C. Principe,et al.  Neural and adaptive systems : fundamentals through simulations , 2000 .

[27]  P J Webros.  BACKPROPAGATION THROUGH TIME: WHAT IT DOES AND HOW TO DO IT , 1990 .

[28]  Gadadhar Sahoo,et al.  Designing an Artificial Neural Network Model for the Prediction of Thromboembolic Stroke , 2004 .

[29]  Erkki Oja,et al.  Time series prediction competition: The CATS benchmark , 2007, Neurocomputing.

[30]  A. Wolf,et al.  Determining Lyapunov exponents from a time series , 1985 .

[31]  Rudolf Sollacher,et al.  Efficient online learning with Spiral Recurrent Neural Networks , 2008, 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence).

[32]  Danilo P. Mandic,et al.  Recurrent Neural Networks for Prediction , 2001 .

[33]  Clive W. J. Granger.  Preface: Some Thoughts on the Future of Forecasting , 2005 .

[34]  Michael C. Mackey,et al.  From Clocks to Chaos , 1988 .

[35]  Acknowledgments , 2006, Molecular and Cellular Endocrinology.

[36]  D. Broomhead,et al.  Robust estimation of tangent maps and Liapunov spectra , 1996 .

[37]  Alistair I. Mees.  Nonlinear Dynamics and Statistics , 2012 .

[38]  V. Alarcon-Aquino,et al.  Single-Step Prediction of Chaotic Time Series Using Wavelet-Networks , 2006, Electronics, Robotics and Automotive Mechanics Conference (CERMA'06).

[39]  Leon O. Chua,et al.  Practical Numerical Algorithms for Chaotic Systems , 1989 .

[40]  Leonard A. Smith.  Disentangling Uncertainty and Error: On the Predictability of Nonlinear Systems , 2001 .

[41]  Paul J. Werbos,et al.  The Roots of Backpropagation: From Ordered Derivatives to Neural Networks and Political Forecasting , 1994 .

[42]  Guilherme De A. Barreto,et al.  Long-term time series prediction with the NARX network: An empirical evaluation , 2008, Neurocomputing.

[43]  Peter Hunter,et al.  Theory of heart , 1991 .

[44]  Henry D. I. Abarbanel,et al.  Challenges in Modeling Nonlinear Systems: A Worked Example , 2001 .

[45]  C. J. Franco,et al.  PREDICCIÓN DE SERIES TEMPORALES USANDO MÁQUINAS DE VECTORES DE SOPORTE TIME SERIES PREDICTION USING SUPPORT VECTOR MACHINES , 2010 .

[46]  Dong-Chul Park,et al.  Multiscale BiLinear Recurrent Neural Networks and Their Application to the Long-Term Prediction of Network Traffic , 2006, ISNN.

[47]  H. Abarbanel,et al.  Prediction in chaotic nonlinear systems: Methods for time series with broadband Fourier spectra. , 1990, Physical review. A, Atomic, molecular, and optical physics.

[48]  F. Takens.  Detecting strange attractors in turbulence , 1981 .

[49]  Sawada,et al.  Measurement of the Lyapunov spectrum from a chaotic time series. , 1985, Physical review letters.

[50]  I SapankevychNicholas,et al.  Time series prediction using support vector machines , 2009 .

[51]  Maria del Pilar Gomez-Gil.  The effect of non-linear dynamic invariants in recurrent neural networks for prediction of electrocardiograms , 1998 .

[52]  HONGXUAN ZHANG,et al.  Complexity Information Based Analysis of Pathological ECG Rhythm for Ventricular Tachycardia and Ventricular Fibrillation , 2002, Int. J. Bifurc. Chaos.