Abstract:The estimation of equations from data is of interest in physics. One of the famous methods is the sparse identification of nonlinear dynamics (SINDy), which utilizes sparse estimation techniques to estimate equations from data. Recently, a method based on the Koopman operator has been developed; the generator extended dynamic mode decomposition (gEDMD) estimates a time evolution generator of dynamical and stochastic systems. However, a naive application of the gEDMD algorithm cannot work well for stochastic differential equations because of the noise effects in the data. Hence, the estimation based on conditional expectation values, in which we approximate the first and second derivatives on each coordinate, is practical. A naive approach is the usage of locally weighted expectations. We show that the naive locally weighted expectation is insufficient because of the nonlinear behavior of the underlying system. For improvement, we apply the clustering method in two ways; one is to reduce the effective number of data, and the other is to capture local information more accurately. We demonstrate the improvement of the proposed method for the double-well potential system with state-dependent noise.

What problem does this paper attempt to address?

The paper is primarily dedicated to improving the accuracy of methods based on the Koopman generator operator in estimating the dynamic equations of stochastic systems, especially when the data is noisy. Specifically, the paper addresses the following key issues: 1. **Problems with existing methods**: The existing generator Extended Dynamic Mode Decomposition (gEDMD) algorithm encounters difficulties when dealing with stochastic differential equations, particularly in the presence of noise in the data. This is because the noise effects can severely impact the estimation results. 2. **Improvement goals**: The goal of the paper is to enhance the accuracy of the gEDMD algorithm in estimating stochastic system equations, especially in the case of noisy data. The authors propose some preprocessing techniques to mitigate the impact of noise on the estimation. 3. **Proposed solutions**: - **Weighted expectations**: Using kernel functions for local weighted averaging to compute conditional expectations. This helps to reduce the impact of noise in the data. - **Clustering methods**: First, representative points are selected through clustering for computing weighted expectations; second, an additional clustering is performed using the Dirichlet Process Mixture Model (DPMM) to more accurately capture the local information of the data. 4. **Study subject**: The paper uses a double-well potential system as a case study, which has state-dependent noise, to well demonstrate the effectiveness of the proposed methods. 5. **Experimental validation**: Numerical experiments are conducted to demonstrate the effectiveness of the proposed methods. The results show that after using the additional clustering step, the estimated drift coefficient function is closer to the true value and can better reproduce the sample trajectory behavior of the original system. In summary, this paper aims to improve the gEDMD algorithm by introducing weighted expectations and clustering methods, enabling it to more accurately estimate the dynamic equations of stochastic systems in noisy datasets.

Improvement of system identification of stochastic systems via Koopman generator and locally weighted expectation

Data-driven Method for Dimension Reduction of Nonlinear Randomly Vibrating Systems

A Data-Driven Method to Identify the Probability Density Expression of Nonlinear System under Gaussian White Noise and Harmonic Excitations

Improving Estimation of the Koopman Operator with Kolmogorov-Smirnov Indicator Functions

Subspace dynamic mode decomposition for stochastic Koopman analysis

Beyond expectations: Residual Dynamic Mode Decomposition and Variance for Stochastic Dynamical Systems

The Koopman Expectation: An Operator Theoretic Method for Efficient Analysis and Optimization of Uncertain Hybrid Dynamical Systems

Data-driven approximation of Koopman operators and generators: Convergence rates and error bounds

Enhancing predictive capabilities in data-driven dynamical modeling with automatic differentiation: Koopman and neural ODE approaches

Modularized data-driven approximation of the Koopman operator and generator

Data-driven probability density forecast for stochastic dynamical systems

A Data-Driven Koopman Approach for Power System Nonlinear Dynamic Observability Analysis

Numerical methods to evaluate Koopman matrix from system equations

Data-driven Fault Diagnosis of Nonlinear Systems With Parameter Uncertainty Using Deep Koopman Operator and Weighted Window Extended Dynamic Mode Decomposition

Physics-informed spectral approximation of Koopman operators

A Data-Driven Approximation of the Koopman Operator: Extending Dynamic Mode Decomposition

Auxiliary Functions as Koopman Observables: Data-Driven Analysis of Dynamical Systems via Polynomial Optimization

Nonlinear Data-Driven Approximation of the Koopman Operator

Koopman Spectral Analysis from Noisy Measurements based on Bayesian Learning and Kalman Smoothing

Learning Koopman-based Stability Certificates for Unknown Nonlinear Systems

Online Real-time Learning of Dynamical Systems from Noisy Streaming Data: A Koopman Operator Approach