Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics
Abstract
We address a three-tier numerical framework based on nonlinear manifold learning for the forecasting of high-dimensional time series, relaxing the “curse of dimensionality” related to the training phase of surrogate/machine learning models. At the first step, we embed the high-dimensional time series into a reduced low-dimensional space using nonlinear manifold learning (local linear embedding and parsimonious diffusion maps). Then, we construct reduced-order surrogate models on the manifold (here, for our illustrations, we used multivariate autoregressive and Gaussian process regression models) to forecast the embedded dynamics. Finally, we solve the pre-image problem, thus lifting the embedded time series back to the original high-dimensional space using radial basis function interpolation and geometric harmonics. The proposed numerical data-driven scheme can also be applied as a reduced-order model procedure for the numerical solution/propagation of the (transient) dynamics of partial differential equations (PDEs). In conclusion, we assess the performance of the proposed scheme via three different families of problems: (a) the forecasting of synthetic time series generated by three simplistic linear and weakly nonlinear stochastic models resembling electroencephalography signals, (b) the prediction/propagation of the solution profiles of a linear parabolic PDE and the Brusselator model (a set of two nonlinear parabolic PDEs), and (c) themore »
- Authors:
-
- University of Naples Federico II (Italy)
- Technion-Israel Institute of Technology, Haifa (Israel)
- Johns Hopkins University, Baltimore, MD (United States)
- Publication Date:
- Research Org.:
- Johns Hopkins University, Baltimore, MD (United States)
- Sponsoring Org.:
- USDOE; US Air Force Office of Scientific Research (AFOSR)
- OSTI Identifier:
- 1982387
- Alternate Identifier(s):
- OSTI ID: 1880176
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Chaos: An Interdisciplinary Journal of Nonlinear Science
- Additional Journal Information:
- Journal Volume: 32; Journal Issue: 8; Journal ID: ISSN 1054-1500
- Publisher:
- American Institute of Physics (AIP)
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS; 97 MATHEMATICS AND COMPUTING; Mathematics; Physics; Dynamical systems; Machine learning; Numerical methods; Gaussian processes; Time series analysis
Citation Formats
Papaioannou, Panagiotis G., Talmon, Ronen, Kevrekidis, Ioannis G., and Siettos, Constantinos. Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics. United States: N. p., 2022.
Web. doi:10.1063/5.0094887.
Papaioannou, Panagiotis G., Talmon, Ronen, Kevrekidis, Ioannis G., & Siettos, Constantinos. Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics. United States. https://doi.org/10.1063/5.0094887
Papaioannou, Panagiotis G., Talmon, Ronen, Kevrekidis, Ioannis G., and Siettos, Constantinos. Mon .
"Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics". United States. https://doi.org/10.1063/5.0094887. https://www.osti.gov/servlets/purl/1982387.
@article{osti_1982387,
title = {Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics},
author = {Papaioannou, Panagiotis G. and Talmon, Ronen and Kevrekidis, Ioannis G. and Siettos, Constantinos},
abstractNote = {We address a three-tier numerical framework based on nonlinear manifold learning for the forecasting of high-dimensional time series, relaxing the “curse of dimensionality” related to the training phase of surrogate/machine learning models. At the first step, we embed the high-dimensional time series into a reduced low-dimensional space using nonlinear manifold learning (local linear embedding and parsimonious diffusion maps). Then, we construct reduced-order surrogate models on the manifold (here, for our illustrations, we used multivariate autoregressive and Gaussian process regression models) to forecast the embedded dynamics. Finally, we solve the pre-image problem, thus lifting the embedded time series back to the original high-dimensional space using radial basis function interpolation and geometric harmonics. The proposed numerical data-driven scheme can also be applied as a reduced-order model procedure for the numerical solution/propagation of the (transient) dynamics of partial differential equations (PDEs). In conclusion, we assess the performance of the proposed scheme via three different families of problems: (a) the forecasting of synthetic time series generated by three simplistic linear and weakly nonlinear stochastic models resembling electroencephalography signals, (b) the prediction/propagation of the solution profiles of a linear parabolic PDE and the Brusselator model (a set of two nonlinear parabolic PDEs), and (c) the forecasting of a real-world data set containing daily time series of ten key foreign exchange rates spanning the time period 3 September 2001–29 October 2020.},
doi = {10.1063/5.0094887},
journal = {Chaos: An Interdisciplinary Journal of Nonlinear Science},
number = 8,
volume = 32,
place = {United States},
year = {Mon Aug 08 00:00:00 EDT 2022},
month = {Mon Aug 08 00:00:00 EDT 2022}
}
Works referenced in this record:
Reduced Models in Chemical Kinetics via Nonlinear Data-Mining
journal, January 2014
- Chiavazzo, Eliodoro; Gear, Charles; Dsilva, Carmeline
- Processes, Vol. 2, Issue 1
Predicting Spatio-temporal Time Series Using Dimension Reduced Local States
journal, October 2019
- Isensee, Jonas; Datseris, George; Parlitz, Ulrich
- Journal of Nonlinear Science, Vol. 30, Issue 3
Time Series Forecasting with Gaussian Processes Needs Priors
book, January 2021
- Corani, Giorgio; Benavoli, Alessio; Zaffalon, Marco
- Machine Learning and Knowledge Discovery in Databases. Applied Data Science Track
A divide and conquer method for the symmetric tridiagonal eigenproblem
journal, June 1980
- Cuppen, J. J. M.
- Numerische Mathematik, Vol. 36, Issue 2
Observing spatio-temporal dynamics of excitable media using reservoir computing
journal, April 2018
- Zimmermann, Roland S.; Parlitz, Ulrich
- Chaos: An Interdisciplinary Journal of Nonlinear Science, Vol. 28, Issue 4
Can social microblogging be used to forecast intraday exchange rates?
journal, October 2013
- Papaioannou, Panagiotis; Russo, Lucia; Papaioannou, George
- NETNOMICS: Economic Research and Electronic Networking, Vol. 14, Issue 1-2
Dynamic mode decomposition of numerical and experimental data
journal, July 2010
- Schmid, Peter J.
- Journal of Fluid Mechanics, Vol. 656
The Isomap Algorithm and Topological Stability
journal, January 2002
- Balasubramanian, M.
- Science, Vol. 295, Issue 5552
On the Koopman Operator of Algorithms
journal, January 2020
- Dietrich, Felix; Thiem, Thomas N.; Kevrekidis, Ioannis G.
- SIAM Journal on Applied Dynamical Systems, Vol. 19, Issue 2
Data-Driven Reduction for a Class of Multiscale Fast-Slow Stochastic Dynamical Systems
journal, January 2016
- Dsilva, Carmeline J.; Talmon, Ronen; Gear, C. William
- SIAM Journal on Applied Dynamical Systems, Vol. 15, Issue 3
Manifold parametrizations by eigenfunctions of the Laplacian and heat kernels
journal, February 2008
- Jones, P. W.; Maggioni, M.; Schul, R.
- Proceedings of the National Academy of Sciences, Vol. 105, Issue 6
Earthquake-explosion discrimination using diffusion maps
journal, September 2016
- Rabin, N.; Bregman, Y.; Lindenbaum, O.
- Geophysical Journal International, Vol. 207, Issue 3
A Global Geometric Framework for Nonlinear Dimensionality Reduction
journal, December 2000
- Tenenbaum, J. B.
- Science, Vol. 290, Issue 5500
To infinity and some glimpses of beyond
journal, November 2017
- Kevrekidis, Panayotis G.; Siettos, Constantinos I.; Kevrekidis, Yannis G.
- Nature Communications, Vol. 8, Issue 1
Nonlinear principal component analysis using autoassociative neural networks
journal, February 1991
- Kramer, Mark A.
- AIChE Journal, Vol. 37, Issue 2
A nonlinear manifold-based reduced order model for multiscale analysis of heterogeneous hyperelastic materials
journal, May 2016
- Bhattacharjee, Satyaki; Matouš, Karel
- Journal of Computational Physics, Vol. 313
Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks
journal, May 2018
- Vlachas, Pantelis R.; Byeon, Wonmin; Wan, Zhong Y.
- Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 474, Issue 2213
ARPACK Users' Guide
book, January 1998
- Lehoucq, R. B.; Sorensen, D. C.; Yang, C.
- Software, Environments, and Tools
Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach
journal, January 2018
- Pathak, Jaideep; Hunt, Brian; Girvan, Michelle
- Physical Review Letters, Vol. 120, Issue 2
The Runge phenomenon and spatially variable shape parameters in RBF interpolation
journal, August 2007
- Fornberg, Bengt; Zuev, Julia
- Computers & Mathematics with Applications, Vol. 54, Issue 3
Diffusion maps
journal, July 2006
- Coifman, Ronald R.; Lafon, Stéphane
- Applied and Computational Harmonic Analysis, Vol. 21, Issue 1
The Gap-Tooth Scheme for Homogenization Problems
journal, January 2005
- Samaey, Giovanni; Roose, Dirk; Kevrekidis, Ioannis G.
- Multiscale Modeling & Simulation, Vol. 4, Issue 1
25 years of time series forecasting
journal, January 2006
- De Gooijer, Jan G.; Hyndman, Rob J.
- International Journal of Forecasting, Vol. 22, Issue 3
LSTM: A Search Space Odyssey
journal, October 2017
- Greff, Klaus; Srivastava, Rupesh K.; Koutnik, Jan
- IEEE Transactions on Neural Networks and Learning Systems, Vol. 28, Issue 10
A comparison of direct and iterated multistep AR methods for forecasting macroeconomic time series
journal, November 2006
- Marcellino, Massimiliano; Stock, James H.; Watson, Mark W.
- Journal of Econometrics, Vol. 135, Issue 1-2
Patch dynamics with buffers for homogenization problems
journal, March 2006
- Samaey, Giovanni; Kevrekidis, Ioannis G.; Roose, Dirk
- Journal of Computational Physics, Vol. 213, Issue 1
datafold: data-driven models for point clouds and time series on manifolds
journal, July 2020
- Lehmberg, Daniel; Dietrich, Felix; Köster, Gerta
- Journal of Open Source Software, Vol. 5, Issue 51
Coarse-grained variables for particle-based models: diffusion maps and animal swarming simulations
journal, September 2014
- Liu, Ping; Safford, Hannah R.; Couzin, Iain D.
- Computational Particle Mechanics, Vol. 1, Issue 4
Fredholm and Volterra Integral Equations of the Second Kind
journal, January 1990
- Press, William H.; Teukolsky, Saul A.
- Computers in Physics, Vol. 4, Issue 5
A data-driven reduced-order model of nonlinear processes based on diffusion maps and artificial neural networks
journal, October 2020
- Koronaki, E. D.; Nikas, A. M.; Boudouvis, A. G.
- Chemical Engineering Journal, Vol. 397
Intrinsic Isometric Manifold Learning with Application to Localization
journal, January 2019
- Schwartz, Ariel; Talmon, Ronen
- SIAM Journal on Imaging Sciences, Vol. 12, Issue 3
Nonlinear Component Analysis as a Kernel Eigenvalue Problem
journal, July 1998
- Schölkopf, Bernhard; Smola, Alexander; Müller, Klaus-Robert
- Neural Computation, Vol. 10, Issue 5
Gaussian processes for time-series modelling
journal, February 2013
- Roberts, S.; Osborne, M.; Ebden, M.
- Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 371, Issue 1984
Time series forecasting for nonlinear and non-stationary processes: a review and comparative study
journal, January 2015
- Cheng, Changqing; Sa-Ngasoongsong, Akkarapol; Beyca, Omer
- IIE Transactions, Vol. 47, Issue 10
Applied Koopman Theory for Partial Differential Equations and Data-Driven Modeling of Spatio-Temporal Systems
journal, December 2018
- Nathan Kutz, J.; Proctor, J. L.; Brunton, S. L.
- Complexity, Vol. 2018
Molecular enhanced sampling with autoencoders: On-the-fly collective variable discovery and accelerated free energy landscape exploration: Molecular Enhanced Sampling with Autoencoders: On-The-Fly Collective Variable Discovery and Accelerated Free Energy Landscape Exploration
journal, September 2018
- Chen, Wei; Ferguson, Andrew L.
- Journal of Computational Chemistry, Vol. 39, Issue 25
Nonlinear Dimensionality Reduction by Locally Linear Embedding
journal, December 2000
- Roweis, S. T.
- Science, Vol. 290, Issue 5500
Inverting nonlinear dimensionality reduction with scale-free radial basis function interpolation
journal, July 2014
- Monnig, Nathan D.; Fornberg, Bengt; Meyer, François G.
- Applied and Computational Harmonic Analysis, Vol. 37, Issue 1
The MATLAB ODE Suite
journal, January 1997
- Shampine, Lawrence F.; Reichelt, Mark W.
- SIAM Journal on Scientific Computing, Vol. 18, Issue 1
"Coarse" stability and bifurcation analysis using time-steppers: A reaction-diffusion example
journal, August 2000
- Theodoropoulos, C.; Qian, Y. -H.; Kevrekidis, I. G.
- Proceedings of the National Academy of Sciences, Vol. 97, Issue 18
Carry Trades and Global Foreign Exchange Volatility
journal, March 2012
- Menkhoff, Lukas; Sarno, Lucio; Schmeling, Maik
- The Journal of Finance, Vol. 67, Issue 2
Reduced-space Gaussian Process Regression for data-driven probabilistic forecast of chaotic dynamical systems
journal, April 2017
- Wan, Zhong Yi; Sapsis, Themistoklis P.
- Physica D: Nonlinear Phenomena, Vol. 345
Partial directed coherence: a new concept in neural structure determination
journal, May 2001
- Baccalá, Luiz A.; Sameshima, Koichi
- Biological Cybernetics, Vol. 84, Issue 6
Parsimonious representation of nonlinear dynamical systems through manifold learning: A chemotaxis case study
journal, May 2018
- Dsilva, Carmeline J.; Talmon, Ronen; Coifman, Ronald R.
- Applied and Computational Harmonic Analysis, Vol. 44, Issue 3
A comparative study of Reservoir Computing strategies for monthly time series prediction
journal, June 2010
- Wyffels, F.; Schrauwen, B.
- Neurocomputing, Vol. 73, Issue 10-12
Emergent Spaces for Coupled Oscillators
journal, May 2020
- Thiem, Thomas N.; Kooshkbaghi, Mahdi; Bertalan, Tom
- Frontiers in Computational Neuroscience, Vol. 14
SciPy 1.0: fundamental algorithms for scientific computing in Python
journal, February 2020
- Virtanen, Pauli; Gommers, Ralf; Oliphant, Travis E.
- Nature Methods
A nonlinear dimensionality reduction framework using smooth geodesics
journal, March 2019
- Gajamannage, Kelum; Paffenroth, Randy; Bollt, Erik M.
- Pattern Recognition, Vol. 87
A Data–Driven Approximation of the Koopman Operator: Extending Dynamic Mode Decomposition
journal, June 2015
- Williams, Matthew O.; Kevrekidis, Ioannis G.; Rowley, Clarence W.
- Journal of Nonlinear Science, Vol. 25, Issue 6
Facing the high-dimensions: Inverse projection with radial basis functions
journal, May 2015
- Amorim, Elisa; Vital Brazil, Emilio; Mena-Chalco, Jesús
- Computers & Graphics, Vol. 48
Dynamic mode decomposition for financial trading strategies
journal, April 2016
- Mann, Jordan; Kutz, J. Nathan
- Quantitative Finance, Vol. 16, Issue 11
Detecting intrinsic slow variables in stochastic dynamical systems by anisotropic diffusion maps
journal, August 2009
- Singer, A.; Erban, R.; Kevrekidis, I. G.
- Proceedings of the National Academy of Sciences, Vol. 106, Issue 38
The Sharpe Ratio
journal, October 1994
- Sharpe, William F.
- The Journal of Portfolio Management, Vol. 21, Issue 1
Coarse-scale PDEs from fine-scale observations via machine learning
journal, January 2020
- Lee, Seungjoon; Kooshkbaghi, Mahdi; Spiliotis, Konstantinos
- Chaos: An Interdisciplinary Journal of Nonlinear Science, Vol. 30, Issue 1
Data Structures for Statistical Computing in Python
conference, January 2010
- McKinney, Wes
- Proceedings of the Python in Science Conference
Investigating Causal Relations by Econometric Models and Cross-spectral Methods
journal, August 1969
- Granger, C. W. J.
- Econometrica, Vol. 37, Issue 3
Convolutional autoencoder and conditional random fields hybrid for predicting spatial-temporal chaos
journal, December 2019
- Herzog, S.; Wörgötter, F.; Parlitz, U.
- Chaos: An Interdisciplinary Journal of Nonlinear Science, Vol. 29, Issue 12
A Limited Memory Algorithm for Bound Constrained Optimization
journal, September 1995
- Byrd, Richard H.; Lu, Peihuang; Nocedal, Jorge
- SIAM Journal on Scientific Computing, Vol. 16, Issue 5
Statsmodels: Econometric and Statistical Modeling with Python
conference, January 2010
- Seabold, Skipper; Perktold, Josef
- Proceedings of the Python in Science Conference
Reservoir computing approaches to recurrent neural network training
journal, August 2009
- Lukoševičius, Mantas; Jaeger, Herbert
- Computer Science Review, Vol. 3, Issue 3
Discovering governing equations from data by sparse identification of nonlinear dynamical systems
journal, March 2016
- Brunton, Steven L.; Proctor, Joshua L.; Kutz, J. Nathan
- Proceedings of the National Academy of Sciences, Vol. 113, Issue 15
Dynamic Mode Decomposition
book, January 2016
- Kutz, J. Nathan; Brunton, Steven L.; Brunton, Bingni W.
- Society for Industrial and Applied Mathematics
Laplacian Eigenmaps for Dimensionality Reduction and Data Representation
journal, June 2003
- Belkin, Mikhail; Niyogi, Partha
- Neural Computation, Vol. 15, Issue 6
Detecting strange attractors in turbulence
book, January 1981
- Takens, Floris
- Lecture Notes in Mathematics
Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization
journal, December 1997
- Zhu, Ciyou; Byrd, Richard H.; Lu, Peihuang
- ACM Transactions on Mathematical Software, Vol. 23, Issue 4
Solving Ill-Posed Linear Systems with GMRES and a Singular Preconditioner
journal, January 2012
- Eldén, Lars; Simoncini, Valeria
- SIAM Journal on Matrix Analysis and Applications, Vol. 33, Issue 4
Equation-free: The computer-aided analysis of complex multiscale systems
journal, January 2004
- Kevrekidis, Ioannis G.; Gear, C. William; Hummer, Gerhard
- AIChE Journal, Vol. 50, Issue 7
Analysis of Fluid Flows via Spectral Properties of the Koopman Operator
journal, January 2013
- Mezić, Igor
- Annual Review of Fluid Mechanics, Vol. 45, Issue 1
Coarse Brownian dynamics for nematic liquid crystals: Bifurcation, projective integration, and control via stochastic simulation
journal, June 2003
- Siettos, C. I.; Graham, M. D.; Kevrekidis, I. G.
- The Journal of Chemical Physics, Vol. 118, Issue 22
Symmetry Breaking Instabilities in Dissipative Systems. II
journal, February 1968
- Prigogine, I.; Lefever, R.
- The Journal of Chemical Physics, Vol. 48, Issue 4
A Flexible Inner-Outer Preconditioned GMRES Algorithm
journal, March 1993
- Saad, Youcef
- SIAM Journal on Scientific Computing, Vol. 14, Issue 2
The gap-tooth method in particle simulations
journal, September 2003
- Gear, C. William; Li, Ju; Kevrekidis, Ioannis G.
- Physics Letters A, Vol. 316, Issue 3-4
Robust kernel Isomap
journal, March 2007
- Choi, Heeyoul; Choi, Seungjin
- Pattern Recognition, Vol. 40, Issue 3
Gaussian Processes in Machine Learning
book, January 2004
- Rasmussen, Carl Edward
- Advanced Lectures on Machine Learning
Equation-Free, Coarse-Grained Multiscale Computation: Enabling Mocroscopic Simulators to Perform System-Level Analysis
journal, January 2003
- Gear, C. William; Hyman, James M.; Kevrekidid, Panagiotis G.
- Communications in Mathematical Sciences, Vol. 1, Issue 4
A Nonlinear Causality Estimator Based on Non-Parametric Multiplicative Regression
journal, June 2016
- Nicolaou, Nicoletta; Constandinou, Timothy G.
- Frontiers in Neuroinformatics, Vol. 10
Geometric diffusions as a tool for harmonic analysis and structure definition of data: Diffusion maps
journal, May 2005
- Coifman, R. R.; Lafon, S.; Lee, A. B.
- Proceedings of the National Academy of Sciences, Vol. 102, Issue 21
Attractor Modeling and Empirical Nonlinear Model Reduction of Dissipative Dynamical Systems
journal, April 2007
- Bollt, Erik
- International Journal of Bifurcation and Chaos, Vol. 17, Issue 04
Geometric harmonics: A novel tool for multiscale out-of-sample extension of empirical functions
journal, July 2006
- Coifman, Ronald R.; Lafon, Stéphane
- Applied and Computational Harmonic Analysis, Vol. 21, Issue 1
Diffusion maps, spectral clustering and reaction coordinates of dynamical systems
journal, July 2006
- Nadler, Boaz; Lafon, Stéphane; Coifman, Ronald R.
- Applied and Computational Harmonic Analysis, Vol. 21, Issue 1, p. 113-127
Diffusion Maps, Reduction Coordinates, and Low Dimensional Representation of Stochastic Systems
journal, January 2008
- Coifman, R. R.; Kevrekidis, I. G.; Lafon, S.
- Multiscale Modeling & Simulation, Vol. 7, Issue 2
Backpropagation algorithms and Reservoir Computing in Recurrent Neural Networks for the forecasting of complex spatiotemporal dynamics
journal, June 2020
- Vlachas, P. R.; Pathak, J.; Hunt, B. R.
- Neural Networks, Vol. 126
Data-driven model reduction, Wiener projections, and the Koopman-Mori-Zwanzig formalism
journal, January 2021
- Lin, Kevin K.; Lu, Fei
- Journal of Computational Physics, Vol. 424
Symmetric Gauge Functions and Unitarily Invariant Norms
journal, January 1960
- Mirsky, L.
- The Quarterly Journal of Mathematics, Vol. 11, Issue 1
Connecting dots: from local covariance to empirical intrinsic geometry and locally linear embedding
journal, January 2019
- Malik, John; Shen, Chao; Wu, Hau-Tieng
- Pure and Applied Analysis, Vol. 1, Issue 4