Abstract
In this paper, we present a data-driven regression approach to identify parametric governing equations from time-series data. Iterative computations are performed for each time stamp to first determine if the governing equations to be recovered are time dependent. The results are then used as input data to extract the parametric equations. A combination of the constrained \(\ell ^1\) and \(\ell ^0+\ell ^2\) optimization problems are used to ensure parsimonious representation of the learned dynamics in the form of parametric differential equations. The method is demonstrated on three canonical dynamics. We show that the proposed method outperforms other sparse-promoting algorithms in identifying parametric differential equations in the low-noise regime in the aspect of accuracy and computation time.
Similar content being viewed by others
Data availability
The data used in this paper can be generated using the method as described in Sect. 4. Sample code is available on https://github.com/HuimeiMa/ParametricDynamicModelSelection.
References
Chou, I.-C., Voit, E.O.: Recent developments in parameter estimation and structure identification of biochemical and genomic systems. Math. Biosci. 219(2), 57–83 (2009)
Engl, H.W., Flamm, C., Kügler, P., Lu, J., Müller, S., Schuster, P.: Inverse problems in systems biology. Inverse Prob. 25(12), 123014 (2009)
Wang, W.-X., Lai, Y.-C., Grebogi, C.: Data based identification and prediction of nonlinear and complex dynamical systems. Phys. Rep. 644, 1–76 (2016)
Bongard, J., Lipson, H.: Automated reverse engineering of nonlinear dynamical systems. Proc. Natl. Acad. Sci. 104(24), 9943–9948 (2007)
Schmidt, M., Lipson, H.: Distilling free-form natural laws from experimental data. Science 324(5923), 81–85 (2009)
Brunton, S.L., Proctor, J.L., Kutz, J.N.: Discovering governing equations from data by sparse identification of nonlinear dynamical systems. Proc. Natl. Acad. Sci. 113(15), 3932–3937 (2016)
Rudy, S.H., Brunton, S.L., Proctor, J.L., Kutz, J.N.: Data-driven discovery of partial differential equations. Sci. Adv. 3(4), 1602614 (2017)
Mangan, N.M., Brunton, S.L., Proctor, J.L., Kutz, J.N.: Inferring biological networks by sparse identification of nonlinear dynamics. IEEE Trans. Mol. Biol. Multi-Scale Commun. 2(1), 52–63 (2016)
Kaheman, K., Kutz, J.N., Brunton, S.L.: SINDy-PI: a robust algorithm for parallel implicit sparse identification of nonlinear dynamics. Proc. R. Soc. A 476(2242), 20200279 (2020)
Brunton, S.L., Proctor, J.L., Kutz, J.N.: Sparse identification of nonlinear dynamics with control (SINDYc). IFAC-PapersOnLine 49(18), 710–715 (2016)
Kaiser, E., Kutz, J.N., Brunton, S.L.: Sparse identification of nonlinear dynamics for model predictive control in the low-data limit. Proc. R. Soc. A 474(2219), 20180335 (2018)
Fasel, U., Kaiser, E., Kutz, J.N., Brunton, B.W., Brunton, S.L.: SINDy with control: A tutorial. In: 2021 60th IEEE Conference on Decision and Control (CDC), pp. 16–21 (2021). IEEE
Shea, D.E., Brunton, S.L., Kutz, J.N.: SINDy-BVP: sparse identification of nonlinear dynamics for boundary value problems. Phys. Rev. Res. 3(2), 023255 (2021)
Schaeffer, H., McCalla, S.G.: Sparse model selection via integral terms. Phys. Rev. E 96(2), 023302 (2017)
Messenger, D.A., Bortz, D.M.: Weak SINDy for partial differential equations. J. Comput. Phys. 443, 110525 (2021)
Messenger, D.A., Bortz, D.M.: Weak SINDy: Galerkin-based data-driven model selection. Multiscale Model. Simul. 19(3), 1474–1497 (2021)
Bortz, D.M., Messenger, D.A., Dukic, V.: Direct estimation of parameters in ODE models using WENDy: weak-form estimation of nonlinear dynamics. arXiv preprint arXiv:2302.13271 (2023)
Schaeffer, H.: Learning partial differential equations via data discovery and sparse optimization. Proc. R. Soc. A Math. Phys. Eng. Sci. 473(2197), 20160446 (2017)
Tibshirani, R.: Regression shrinkage and selection via the lasso. J. R. Stat. Soc. Ser. B (Methodol.) 58(1), 267–288 (1996)
Lions, P.-L., Mercier, B.: Splitting algorithms for the sum of two nonlinear operators. SIAM J. Numer. Anal. 16(6), 964–979 (1979)
Combettes, P.L., Pesquet, J.-C.: Proximal splitting methods in signal processing. Fixed-point Algorithms Inverse Probl. Sci. Eng. 185–212 (2011)
Combettes, P.L., Pesquet, J.-C.: A Douglas-Rachford splitting approach to nonsmooth convex variational signal recovery. IEEE J. Sel. Top. Signal Process. 1(4), 564–574 (2007)
He, B., Yuan, X.: On the \({\cal{O} }(1/n)\)convergence rate of the Douglas-Rachford alternating direction method. SIAM J. Numer. Anal. 50(2), 700–709 (2012)
Schaeffer, H., Tran, G., Ward, R.: Extracting sparse high-dimensional dynamics from limited data. SIAM J. Appl. Math. 78(6), 3279–3295 (2018)
Schaeffer, H., Tran, G., Ward, R., Zhang, L.: Extracting structured dynamical systems using sparse optimization with very few samples. Multiscale Model. Simul. 18(4), 1435–1461 (2020)
Eckstein, J.: Splitting Methods for Monotone Operators with Applications to Parallel Optimization. PhD thesis, Massachusetts Institute of Technology (1989)
Foucart, S., Rauhut, H.: A Mathematical Introduction to Compressive Sensing. Birkhäuser, New York (2013)
Gabay, D., Mercier, B.: A dual algorithm for the solution of nonlinear variational problems via finite element approximation. Comput. Math. Appl. 2(1), 17–40 (1976)
Boyd, S., Parikh, N., Chu, E., Peleato, B., Eckstein, J., et al.: Distributed optimization and statistical learning via the alternating direction method of multipliers. Found. Trends® Mach. Learn. 3(1), 1–122 (2011)
Daubechies, I., Defrise, M., De Mol, C.: An iterative thresholding algorithm for linear inverse problems with a sparsity constraint. Commun. Pure Appl. Math. 57(11), 1413–1457 (2004)
Beck, A., Teboulle, M.: A fast iterative shrinkage-thresholding algorithm for linear inverse problems. SIAM J. Imag. Sci. 2(1), 183–202 (2009)
Van Den Berg, E., Friedlander, M.P.: Probing the Pareto frontier for basis pursuit solutions. SIAM J. Sci. Comput. 31(2), 890–912 (2009)
Chambolle, A., Pock, T.: A first-order primal-dual algorithm for convex problems with applications to imaging. J. Math. Imaging Vis. 40, 120–145 (2011)
Rudy, S., Alla, A., Brunton, S.L., Kutz, J.N.: Data-driven identification of parametric partial differential equations. SIAM J. Appl. Dyn. Syst. 18(2), 643–660 (2019)
Li, X., Li, L., Yue, Z., Tang, X., Voss, H., Kurths, J., Yuan, Y.: Sparse learning of partial differential equations with structured dictionary matrix. Chaos Interdiscip. J. Nonlinear Sci. 29, 043130 (2019). https://doi.org/10.1063/1.5054708
Grant, M., Boyd, S.: CVX: Matlab Software for Disciplined Convex Programming, version 2.1. http://cvxr.com/cvx (2014)
Grant, M.C., Boyd, S.P.: Graph implementations for nonsmooth convex programs. In: Recent Advances in Learning and Control, pp. 95–110 (2008). Springer, New York
Li, Z., Kovachki, N., Azizzadenesheli, K., Liu, B., Bhattacharya, K., Stuart, A., Anandkumar, A.: Fourier neural operator for parametric partial differential equations. arXiv preprint arXiv:2010.08895 (2020)
Xu, H., Zhang, D., Zeng, J.: Deep-learning of parametric partial differential equations from sparse and noisy data. Phys. Fluids 33(3), 037132 (2021)
Wang, S., Wang, H., Perdikaris, P.: Learning the solution operator of parametric partial differential equations with physics-informed deeponets. Sci. Adv. 7(40), 8605 (2021)
Im, J., Rizzo, C.B., Barros, F.P., Masri, S.F.: Application of genetic programming for model-free identification of nonlinear multi-physics systems. Nonlinear Dyn. 104, 1781–1800 (2021)
Chen, Y., Luo, Y., Liu, Q., Xu, H., Zhang, D.: Symbolic genetic algorithm for discovering open-form partial differential equations (SGA-PDE). Phys. Rev. Res. 4(2), 023174 (2022)
Davis, D., Yin, W.: Convergence rate analysis of several splitting schemes. Split. Methods Commun. Imaging Sci. Eng. 115–163 (2016)
Giselsson, P., Boyd, S.: Linear convergence and metric selection for Douglas-Rachford splitting and ADMM. IEEE Trans. Autom. Control 62(2), 532–544 (2016)
Fisher, R.A.: The wave of advance of advantageous genes. Ann. Eugen. 7(4), 355–369 (1937)
Tikhomirov, V.M.: A study of the diffusion equation with increase in the amount of substance, and its application to a biological problem, pp. 242–270. Springer, New York (1991)
Reinbold, P.A.K., Gurevich, D.R., Grigoriev, R.O.: Using noisy or incomplete data to discover models of spatiotemporal dynamics. Phys. Rev. E 101(1), 010203 (2020)
Kaptanoglu, A.A., Silva, B.M., Fasel, U., Kaheman, K., Goldschmidt, A.J., Callaham, J.L., Delahunt, C.B., Nicolaou, Z.G., Champion, K., Loiseau, J.-C., et al.: Pysindy: A comprehensive python package for robust sparse system identification. arXiv preprint arXiv:2111.08481 (2021)
Antonelli, G., Chiaverini, S., Di Lillo, P.: On data-driven identification: Is automatically discovering equations of motion from data a chimera? Nonlinear Dyn. 111(7), 6487–6498 (2023)
Strebel, O.: Preprocessing algorithms for the estimation of ordinary differential equation models with polynomial nonlinearities. Nonlinear Dyn. 1–16 (2023)
Mangan, N.M., Kutz, J.N., Brunton, S.L., Proctor, J.L.: Model selection for dynamical systems via sparse regression and information criteria. Proc. R. Soc. A Math. Phys. Eng. Sci. 473(2204), 20170009 (2017)
Baake, E., Baake, M., Bock, H.G., Briggs, K.M.: Fitting ordinary differential equations to chaotic data. Phys. Rev. A 45(8), 5524 (1992)
Douglas, J., Rachford, H.H.: On the numerical solution of heat conduction problems in two and three space variables. Trans. Am. Math. Soc. 82(2), 421–439 (1956)
Patrinos, P., Stella, L., Bemporad, A.: Douglas-Rachford splitting: Complexity estimates and accelerated variants. In: 53rd IEEE Conference on Decision and Control, pp. 4234–4239 (2014). IEEE
Pham, M., Rana, A., Miao, J., Osher, S.: Semi-implicit relaxed Douglas-Rachford algorithm (sDR) for ptychography. Opt. Express 27(22), 31246–31260 (2019)
Fu, A., Zhang, J., Boyd, S.: Anderson accelerated Douglas-Rachford splitting. SIAM J. Sci. Comput. 42(6), 3560–3583 (2020)
Goldstein, T., Osher, S.: The split Bregman method for \(l^1\)-regularized problems. SIAM J. Imag. Sci. 2(2), 323–343 (2009)
Acknowledgements
L. Zhang was supported by NSFC Grant #12101342.
Funding
The authors have not disclosed any funding.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: The Douglas-Rachford algorithm
Consider the minimization problem (1.2), where \(G_1\) and \(G_2\) are functions for which one can compute the proximal mappings \(\text {prox}_{\gamma G_i}\), \(i=1,2\), via Eq. (1.3). The DR algorithm was introduced in [20] as a generalization of an algorithm introduced by Douglas and Rachford in the case of quadratic minimization [53]. Under certain conditions, the DR algorithm have the following convergence property [22, 23, 43, 44].
Theorem A.1
Let \(G_1\) and \(G_2\) be proper, closed, and convex functions. For any \(\gamma >0\), any \(\mu \in (0,2)\), and any initial point \(\tilde{x}^0\), the iterates \(x^k\) generated by Eq. (1.4) converges linearly to a minimizer of the minimization problem (1.2).
For the constrained \(\ell ^1\) minimization problem (P\(_{1,\epsilon }\)) of main interest in this paper, we define
where
and \(\text {Ind}\) denotes the indicator function. The proximal operators \(\text {prox}_{\gamma G_i}\), \(i=1,2\), are then given by:
where
The function S is the soft-thresholding function which is defined component-wise as follows:
and \(\text {prox}_{\mathcal {B}}\) is the projection operator onto the ball \(\mathcal {B}\):
For the constrained \(\ell ^0\) minimization problem (P\(_{0,\epsilon }\)) considered for comparison in Sect. 5, redefine \(G_1\) as follows:
The proximal operator \(\text {prox}_{\gamma {\tilde{G}}_1}\) is given by:
where the function H is the hard-thresholding function which is defined component-wise as follows:
In Sect. 3, we use the basic form of the DR algorithm, i.e. Eq. (1.4). In [54,55,56], accelerated variants of the DR algorithm are presented. To solve Eq. (P\(_{1,\epsilon }\)), one can also choose other algorithms for \(\ell ^1\) minimization [27], for example ADMM, SPGL1, and the split Bregman algorithm [57].
Appendix B: Hyperparameters and supplementary figures
In this section, we provide the hyperparameters and some supplementary figures related to the computational experiments in Sect. 4.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ma, H., Lu, X. & Zhang, L. Extracting parametric dynamics from time-series data. Nonlinear Dyn 111, 15177–15199 (2023). https://doi.org/10.1007/s11071-023-08643-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-023-08643-z