Tensor-based computation of the Koopman generator via operator logarithm

Tatsuya Kishimoto; Jun Ohkubo

cs.LGApr 9, 2026

Tensor-based computation of the Koopman generator via operator logarithm

Authors: Tatsuya Kishimoto, Jun Ohkubo

Organizations: Graduate School of Science and Engineering, Saitama University 255 Shimo-Okubo, Sakura-ku, Saitama 338–8570, Japan

Abstract

Identifying governing equations of nonlinear dynamical systems from data is challenging. While sparse identification of nonlinear dynamics (SINDy) and its extensions are widely used for system identification, operator-logarithm approaches use the logarithm to avoid time differentiation, enabling larger sampling intervals. However, they still suffer from the curse of dimensionality. Then, we propose a data-driven method to compute the Koopman generator in a low-rank tensor train (TT) format by taking logarithms of Koopman eigenvalues while preserving the TT format. Experiments on 4-dimensional Lotka-Volterra and 10-dimensional Lorenz-96 systems show accurate recovery of vector field coefficients and scalability to higher-dimensional systems.

Figures & tables

Figure 1: Segment of the vdP oscillator trajectory data used for EDMD. The trajectory is generated from the initial condition

[x_{1},x_{2}]=[1.0,1.0]

, and this plot shows the part from

t=0

t=10

sampled with a sampling interval of

T_{\mathrm{s}}=0.1

Figure 2: Illustration of the proposed method for computing the Koopman generator in the TT format. (a) shows the conventional matrix-logarithm-based approach, whereas (b) summarizes the proposed approach based on the eigendecomposition relationship. In the proposed method, the logarithm is applied to the eigenvalues rather than to the Koopman matrix, enabling efficient computation in high-dimensional settings. The Koopman matrix and its eigendecomposition are shown in parentheses to indicate that the eigenvalues and eigentensors can be computed directly using AMUSEt.

Figure 3: Tensor contraction process for computing the Koopman generator in the TT format.

\mathbf{\Xi}

is the Koopman eigentensor,

\Sigma

is the diagonal matrix of the eigenvalues of the Koopman generator, and

\mathbf{\Xi}^{+}

is the pseudo-inverse of the Koopman eigentensor. By contracting the last mode of

\mathbf{\Sigma}

and the first mode of

\mathbf{\Xi}^{+}

, we can obtain the Koopman generator

\mathbf{L}

in the TT format.

\mathbf{\Xi}^{+}=\left\{\mathbf{U}_{X,r}\Sigma^{-1}_{X,r}\left\{A^{+}\right\}^{\mathsf{T}}\right\}^{\mathsf{T}}.

Algorithm 1 Proposed method for Koopman generator identification in the TT format

Figure 4: Vector field coefficients obtained for the Lotka–Volterra system. The horizontal axis represents the indices of the coefficient vector, and the vertical axis represents the values of the coefficient vector. The blue dots represent the true coefficients, and the orange dots represent the estimated coefficients. All coefficients are nearly identical between the true and estimated values.

Term	Absolute error
$x_{2}x_{10}$	$3.67\times 10^{-8}$
$x_{3}x_{1}$	$5.18\times 10^{-6}$
$x_{4}x_{2}$	$4.69\times 10^{-8}$
$x_{5}x_{3}$	$6.75\times 10^{-7}$
$x_{6}x_{4}$	$1.56\times 10^{-7}$
$x_{7}x_{5}$	$5.78\times 10^{-8}$

Table 1: Absolute errors in the estimated coefficients of the

x_{i+1}x_{i-1}

terms in the Lorenz–96 system. The absolute error for each coefficient is defined as the absolute value of the difference between the true and estimated values, i.e.,

|\hat{w}_{k,i}-1|

Term	Absolute error
$x_{1}x_{2}$	$9.46\times 10^{-7}$
$x_{2}x_{3}$	$4.22\times 10^{-6}$
$x_{3}x_{4}$	$1.62\times 10^{-6}$
$x_{4}x_{5}$	$9.29\times 10^{-7}$
$x_{5}x_{6}$	$1.19\times 10^{-7}$
$x_{6}x_{7}$	$4.37\times 10^{-7}$

Table 2: Absolute errors in the estimated coefficients of the

x_{i}x_{i+1}

terms, which are not present in the Lorenz–96 system. The absolute error for each coefficient equals the absolute value of the estimate, since the true value is zero.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Figure 5: Graphical representation of the reduced-matrix computation. The tensors

\mathbf{U}_{X,r}

and

\mathbf{U}_{Y,r}

are left-orthonormalized so that the contractions can be performed efficiently. Since we define

\mathbf{\Psi}(Y)=K\mathbf{\Psi}(X)

, the tensor contraction sequence differs from that in the original algorithm.

Explore similar work

Jul 2, 2026eess.SY

Koopman operator theory: fundamentals, control, and applications

The Koopman operator has gained considerable attention due to its ability to provide a global linear representation of highly complex dynamical systems. The operator describes nonlinear dynamics in a linear way through the lens of real- or complex-valued observable functions. Recently proposed data-driven techniques, like extended dynamic mode decomposition (EDMD), its kernelized variant, and machine-learning methods, can be used to generate finite-dimensional approximations accompanied by finite-data error bounds. In this tutorial paper, we provide a concise introduction into Koopman operator theory and its use in systems and control. A particular focus is put on data-driven surrogate models, their extension to systems with inputs, and controller design using Koopman operator theory. Moreover, we demonstrate the key techniques, i.e., EDMD and Koopman MPC. To this end, we provide simulation studies including source code on GitHub to enable the interested reader to experience the Koopman operator in systems and control step by step.

Igor Mezić, Jorge Cortés, Karl Worthmann +2

Department of Mechanical Engineering, University of California, Santa Barbara, USA · Department of Mechanical and Aerospace Engineering, University of California, San Diego, USA · Optimization-based Control Group, TU Ilmenau, Germany +2

Koopman Operator Dynamic Mode Decomposition Robust Control Model Predictive Control Surrogate Models

Ask PDF

May 3, 2026cs.LG

Learning Koopman operators for coupled systems via information on governing equations of subsystems

Nonlinear coupled systems are ubiquitous in science and engineering. The analysis and modeling of such systems is challenging due to their high dimensionality and complex interactions among subsystems. In recent years, operator-theoretic methods based on the Koopman operator have attracted attention as a powerful tool for analyzing and modeling nonlinear dynamical systems. Extended dynamic mode decomposition (EDMD) is one of the most popular methods to approximate the Koopman operator. However, EDMD is a purely data-driven method, and it could be unstable and inaccurate for coupled systems under limited data availability. In this paper, we propose a method to learn the Koopman operator for coupled systems using the differential equations governing each subsystem. We also demonstrate its effectiveness through numerical experiments on coupled oscillator systems.

Tatsuya Naoi, Jun Ohkubo

Graduate School of Science and Engineering, Saitama University, Sakura, Saitama, 338-8570, Japan

Koopman Operator Dynamic Mode Decomposition Nonlinear Dynamics

Ask PDF

Sep 8, 2026cs.LG

Tensor-Train Weak SINDy: Identifying High-Dimensional Nonlinear Dynamics

Weak Sparse Identification of Nonlinear Dynamics (WSINDy) provides a noise-robust approach for learning dynamical systems from data without requiring numerical differentiation. However, for high-dimensional systems, tensor-product libraries of candidate functions grow exponentially with the state dimension, making standard WSINDy expensive in both computation and memory. The Multidimensional Approximation of Nonlinear Dynamics (MANDy) addresses this scaling through a tensor-train (TT) representation of the candidate library, but does not provide a mechanism for sparse model selection. Here, we combine these approaches to develop TT-WSINDy, which performs the weak-form transformation, regression, and sparsification in TT format. We show that the TT formulation recovers the corresponding WSINDy regression problem and derive polynomial time and memory complexity bounds for the tensor-train sparsification procedure. Numerical experiments demonstrate robustness to measurement noise and computational savings for high-dimensional systems.

Will Houser, Vanja Dukic, David M. Bortz

Department of Applied Mathematics, University of Colorado, Boulder, CO 80309-0526 USA

System Identification Nonlinear Dynamics Wavelets

Ask PDF