Abstract
This paper derives error bounds for regression in continuous time over subsets of certain types of Riemannian manifolds. The regression problem is typically driven by a nonlinear evolution law taking values on the manifold, and it is cast as one of optimal estimation in a reproducing kernel Hilbert space. A new notion of persistency of excitation (PE) is defined for the estimation problem over the manifold, and rates of convergence of the continuous time estimates are derived using the PE condition. We discuss and analyze two approximation methods of the exact regression solution. We then conclude the paper with some numerical simulations that illustrate the qualitative character of the computed function estimates. Examples of function estimates generated over a trajectory of the Lorenz system and based on experimental motion capture data are included.
Similar content being viewed by others
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request
References
Bai, S., Wang, J., Chen, F., Englot, B.: Information-theoretic exploration with bayesian optimization. In: 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 1816–1822. IEEE (2016)
Berlinet, A., Thomas-Agnan, C.: Reproducing kernel Hilbert spaces in probability and statistics. Springer Science & Business Media (2011)
Burns, J., Estes, B., Guo, J., Kurdila, A.J., Liu, R., Paruchuri, S.T., Powell, N.: Approximation of koopman operators: domain exploration. Submitted to the 2022 CDC (2022)
Ciarlet, P.G.: Linear and nonlinear functional analysis with applications, vol. 130. Siam (2013)
Cucker, F., Zhou, D.X.: Learning Theory: An Approximation Theory Viewpoint. Cambridge Press (2007)
DeVore, R., Kerkyacharian, G., Picard, D., Temlyakov, V.: Mathematical methods for supervised learning. IMI Preprints 22, 1–51 (2004)
DeVore, R.A., Lorentz, G.G.: Constructive approximation, vol. 303. Springer (1993)
Engl, H.W., Hanke, M., Neubauer, A.: Regularization of inverse problems, vol. 375. Springer Science & Business Media (1996)
Farrell, J.A., Polycarpou, M.M.: Adaptive approximation based control: unifying neural, fuzzy and traditional adaptive approximation approaches, vol. 48. Wiley (2006)
Foster, D., Sarkar, T., Rakhlin, A.: Learning nonlinear dynamical systems from a single trajectory. Learning for Dynamics and Control, PMLR (2020)
Fukuchi, C.A., Fukuchi, R.K., Duarte, M.: A public dataset of overground and treadmill walking kinematics and kinetics in healthy individuals. PeerJ 6, e4640 (2018)
Fuselier, E., Wright, G.B.: Scattered data interpolation on embedded submanifolds with restricted positive definite kernels: Sobolev error estimates. SIAM J. Numer. Anal. 50(3), 1753–1776 (2012)
Gao, T., Kovalsky, S.Z., Daubechies, I.: Gaussian process landmarking on manifolds. SIAM J. Math. Data Sci. 1(1), 208–236 (2019)
Guo, J., Kepler, M.E., Paruchuri, S.T., Wang, H., Kurdila, A.J., Stilwell, D.J.: Strictly decentralized adaptive estimation of external fields using reproducing kernels. arXiv preprint arXiv:2103.12721 (2021)
Guo, J., Paruchuri, S.T., Kurdila, A.J.: Approximations of the reproducing kernel hilbert space (rkhs) embedding method over manifolds. In: 2020 59th IEEE Conference on Decision and Control (CDC), pp. 1596–1601. IEEE (2020)
Guo, J., Paruchuri, S.T., Kurdila, A.J.: Persistence of excitation in uniformly embedded reproducing kernel hilbert (rkh) spaces. In: 2020 American Control Conference (ACC), pp. 4539–4544. IEEE (2020)
Gyorfi, L., Kohler, M., Krzyzak, A., Walk, H.: A Distribution-Free Theory of Nonparametric Regression. Springer (2002)
Hangelbroek, T., Narcowich, F.J., Ward, J.D.: Polyharmonic and related kernels on manifolds: interpolation and approximation. Found. Comput. Math. 12(5), 625–670 (2012)
Hovakimyan, N., Cao, C.: \(\cal{L}_1\) Adaptive Control Theory: Guaranteed Robustness with Fast Adaptation. SIAM (2010)
Ioannou, P.A., Sun, J.: Robust adaptive control. Courier Corporation (2012)
Krstic, M., Kanellakopoulos, I., Kokotovic, P.: Nonlinear and Adaptive Control Design. Wiley (1995)
Kurdila, A.J., Guo, J., Paruchuri, S.T., Bobade, P.: Persistence of excitation in reproducing kernel hilbert spaces, positive limit sets, and smooth manifolds. arXiv preprint arXiv:1909.12274 (2019)
Liu, G.H., Theodorou, E.A.: Deep learning theory review: An optimal control and dynamical systems perspective. arXiv preprint arXiv:1908.10920 (2019)
Narendra, K.S., Annaswamy, A.M.: Stable Adaptive Systems. Dover (1989)
Paruchuri, S.T., Guo, J., Kurdila, A.: Kernel center adaptation in the reproducing kernel hilbert space embedding method. arXiv preprint arXiv:2009.02867 (2020)
Paruchuri, S.T., Guo, J., Kurdila, A.: Sufficient conditions for parameter convergence over embedded manifolds using kernel techniques. arXiv preprint arXiv:2009.02866 (2020)
Paruchuri, S.T., Guo, J., Kurdila, A.: Kernel center adaptation in the reproducing kernel hilbert space embedding method. Int. J. Adapt. Control Signal Process. 36(7), 1562–1583 (2022)
Paruchuri, S.T., Guo, J., Kurdila, A.J.: Sufficient conditions for parameter convergence over embedded manifolds using kernel techniques. IEEE Trans. Autom. Control 68(2), 753–765 (2022)
Pietsch, A.: Approximation spaces. J. Approx. Theory 32(2), 115–134 (1981)
Powell, N., Kurdila, A.J.: Learning theory for estimation of animal motion submanifolds. In: 2020 59th IEEE Conference on Decision and Control (CDC), pp. 4941–4946. IEEE (2020)
Powell, N., Kurdila, A.J.: Distribution-free learning theory for approximating submanifolds from reptile motion capture data. Comput. Mech. 68(2), 337–356 (2021)
Powell, N., Liu, B., Kurdila, A.J.: Koopman methods for estimation of animal motions over unknown, regularly embedded submanifolds. arXiv preprint arXiv:2203.05646 (2022)
Qianxiao, L., Weinan, E.: Machine learning and dynamical systems. SIAM News pp. 5–7 (2021)
Rosasco, L., Belkin, M., Vito, E.D.: On learning with integral operators. J. Mach. Learn. Res. 11, 905–934 (2010)
Sastry, S., Bodson, M.: Adaptive control: stability, convergence and robustness. Courier Corporation (2011)
Smola, B.S.A.J.: Learning with Kernels: Support Vector Machines. Optimization, and Beyond. MIT Press, Regularization (2002)
Temlyakov, V.: Multivariate approximation, vol. 32. Cambridge University Press (2018)
Vapnik, V.N.: An overview of statistical learning theory. IEEE Trans. Neural Netw. 10(5), 988–999 (1999)
Vito, E.D., Rosasco, L., Toigo, A.: Learning sets with separating kernels. Applied and Computational Harmonic Analysis pp. 185–217 (2014)
Walker, J.: Dynamical Systems and Evolution Equations: Theory and Applications. Springer (2013)
Wendland, H.: Scattered data approximation, vol. 17. Cambridge University Press (2004)
Williams, C.K., Rasmussen, C.: Gaussian processes for machine learning. MIT Press (2006)
Wittwar, D.W., Santin, G., Haasdonk, B.: Interpolation with uncoupled separable matrix-valued kernels. arXiv preprint arXiv:1807.09111 (2018)
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
1.1 Background on Galerkin approximations
Let U be a real Hilbert space, \(A\in {\mathcal {L}}(U)\) be a bounded linear operator on U, \(b\in U\) be a fixed element of U, and suppose we seek to find \(u\in U\) that satisfies the operator equation
It is customary that the existence and uniqueness of the solution of this equation is established by studying the associated bilinear form \(a(\cdot ,\cdot ):U\times U \rightarrow \mathbb {R}\) given by \(a(u,v):=(Au,v)\) for all \(u,v\in U\). Then, the operator equation above is equivalent to finding the \(u\in U\) for which
The Lax–Milgram theorem given below stipulates a concise pair of conditions that ensure the well-posedness of the operator equation.
Theorem 3
(Lax–Milgram theorem [4]) The bilinear form \(a(\cdot ,\cdot ):U\times U \rightarrow \mathbb {R}\) is bounded if there is a constant \(C_1>0\) such that
and it is coercive if there is a constant \(C_2>0\) such that
If the bilinear form \(a(\cdot ,\cdot )\) is bounded and coercive, then \(A^{-1}\in {\mathcal {L}}(U)\) and there is a unique solution \(u\in U\) to Eq. (19).
Proof
The first condition above, continuity of the bilinear form, ensures that \(A\in {\mathcal {L}}(U)\) by definition. The coercivity condition implies that the nullspace of A is just \(\{0\}\). As a result, we know that A is one-to-one. This means that the operator \(A^{-1}:\text {range}(A)\rightarrow U\) is well defined. From the coercivity condition, we also conclude that
for every \(b\in \text {range}(A)\). This means that \(A^{-1}\in {\mathcal {L}}(\text {range}(A),U)\) and \(\Vert A^{-1}\Vert \le 1/C_2\).
One implication of the fact that \(A^{-1}\!\in \! {\mathcal {L}}(\text {range}(A),U)\) is that \(\text {range}(A)\) is closed. Suppose that \(\{b_k\}_{k\in \mathbb {N}}\subset \text {range}(A)\) and \(b_k\rightarrow {\bar{b}}\). By construction there is a sequence \(\{u_k\}_{k\in \mathbb {N}}\subset U\) such that \(Au_k=b_k\). But we have
and \(\{u_k\}_{k\in \mathbb {N}}\) is a Cauchy sequence in the complete space U. There is a limit \(u_k\rightarrow {\bar{u}}\in U\). By the continuity of the operator A, we know that \(A{\bar{u}}={\bar{b}}\), hence \({\bar{b}}\in \text {range}(A)\). The range of A is consequently closed.
It only remains to show that \(\text {range}(A)=U\). Suppose to the contrary there is a \({\bar{b}} \ne 0\) with \({\bar{b}} \in (\overline{\text {range}(A)})^\perp \). Since \(\mathcal {N}(A^*) = (\overline{\text {range}(A)})^\perp \), we know that
By the coercivity condition, we must have \(0=\langle A{\bar{b}},{\bar{b}}\rangle _U\ge C_2 \Vert {\bar{b}}\Vert _U^2 \not = 0\). But this is a contradiction and \(\text {range}(A)\) is all of U. \(\square \)
Next, we discuss how error bounds are derived for Galerkin approximations \(u_N\) of the solution u of the operator equations above. Let \(U_N\subseteq U\) be a finite-dimensional subspace of U. By definition, the Galerkin approximation \(u_N\in U_N\) is the unique solution of the equation
The theorem below summarizes one of the well-known bounds on the error \(u-u_N\) between the Galerkin approximation \(u_N\in U_N\) and the true solution \(u\in U\).
Theorem 4
(Cea’s Lemma, [4]) Suppose that the hypotheses of the Lax–Milgram Theorem 3 hold. There is a unique solution \(u_N\in U_N\) of the Galerkin Equation 20. The error \(u-u_N\) is a-orthogonal to the subspace \(U_N\) in the sense that
We also have the error bound
where \(\varPi _N\) is the U-orthogonal projection of U onto \(U_N\).
Proof
First, note that when \(a(\cdot ,\cdot )\) satisfies the boundedness and coercivity conditions, its restriction \(a:U_N\times U_N \rightarrow \mathbb {R}\) to \(U_N\) satisfies the boundedness and coercivity conditions with the same constants relative to \(U_N\). This means that the Galerkin equations have a unique solution \(u_N\in U_N\). Since Eq. (19) holds for all \(v\in U\), it holds for all \(v_N\in U_N\). We can subtract Eqs. 19 and 20 for each \(v_N\in U_N\) and obtain
for each \(v_N\in U_N\). Using the boundedness and coercivity of the bilinear form, as well as the \(a-\)orthogonality of the error, we have
for any \(v_N\in U_N\). The theorem now follows after canceling the common term on the right and left. \(\square \)
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
Burns, J., Estes, B., Guo, J. et al. Kernel methods for regression in continuous time over subsets and manifolds. Nonlinear Dyn 111, 13165–13186 (2023). https://doi.org/10.1007/s11071-023-08567-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-023-08567-8