Abstract:Revealing hidden dynamics from the stochastic data is a challenging problem as the randomness takes part in the evolution of the data. The problem becomes exceedingly hard if the trajectories of the stochastic data are absent in many scenarios. In this work, we propose the Weak Collocation Regression (WCR) method to learn the dynamics from the stochastic data without the labels of trajectories. This method utilize the governing equation of the probability distribution function–the Fokker-Planck (FP) equation. Using its weak form and integration by parts, we move all the spacial derivatives of the distribution function to the test functions which can be computed explicitly. Since the data is a sampling of the corresponding distribution function, we can compute the integrations in the weak form, which has no spacial derivatives on the distribution functions, by simply adding the values of the integrands at the data points. We further assume the unknown drift and diffusion terms can be expanded by the base functions in a dictionary with the coefficients to be determined. Cooperating the collocation treatment and linear multi-step methods, we transfer the revealing process to a linear algebraic system. Using the sparse regression, we eventually obtain the unknown coefficients and hence the hidden stochastic dynamics. The numerical experiments show that our method is flexible and fast, which reveals the dynamics within seconds in the multi-dimensional problems and can be extended to high dimensional data. The complex tasks with variable-dependent diffusion and coupled drift can be correctly identified by WCR and the performance is robust, achieving high accuracy in the cases of noisy data. The rigorous error estimate is also included to support our various numerical experiments.

Stability and Dispersion Analysis of Reproducing Kernel Collocation Method for Transient Dynamics

Stability of Radial Basis Collocation Method for Transient Dynamics

A Collocation Method Based on Reproducing Kernel for A Modified Anomalous Subdiffusion Equation

A New Reproducing Kernel Method for Duffing Equations.

A Subdomain Collocation Method Based on Voronoi Domain Partition and Reproducing Kernel Approximation

Exponential Collocation Methods for Conservative or Dissipative Systems.

Stability Analysis of Discrete-Time Systems with Arbitrary Delay Kernels Based on Kernel-Related Summation Inequality and Model Transformation

Proving the stability estimates of variational least-squares Kernel-Based methods

Time-dependent nonlinear collocation method and stability analysis for natural convection problems

A reproducing kernel based collocation method for arbitrary m-order BVPs

A spectral collocation method for nonlocal diffusion equations

Solving a linear fractional equation with nonlocal boundary conditions based on multiscale orthonormal bases method in the reproducing kernel space

Weak collocation regression method: Fast reveal hidden stochastic dynamics from high-dimensional aggregate data

A POD based reduced-order local RBF collocation approach for time-dependent nonlocal diffusion problems

Convergence and stability of spectral collocation method for hyperbolic partial differential equation with piecewise continuous arguments

Error analysis of a collocation method on graded meshes for nonlocal diffusion problems with weakly singular kernels

Stability of collocation methods for delay differential equations with vanishing delays

Stochastic Collocation Method for Statistical Static Timing Analysis with Input Truncation Technique

Study on point collocated meshless method based on reproducing kernel ideas for one-dimensional problems

Dynamical Effects of Nonlocal Interactions in Discrete-Time Growth-Dispersal Models with Logistic-Type Nonlinearities