A Data-Driven Framework for Discovering Fractional Differential Equations in Complex Systems

TL;DR

This paper presents a novel data-driven framework that uses neural networks and optimization techniques to discover fractional differential equations from noisy data, effectively modeling non-local and memory effects in complex systems.

Contribution

It introduces a stepwise framework combining deep learning, sparse regression, and global optimization to identify fractional differential equations directly from data.

Findings

Robustly identifies FDE structures across various noise levels.

Successfully models systems with long-range interactions and memory effects.

Capable of capturing both fractional and integer-order dynamics.

Abstract

In complex physical systems, conventional differential equations often fall short in capturing non-local and memory effects, as they are limited to local dynamics and integer-order interactions. This study introduces a stepwise data-driven framework for discovering fractional differential equations (FDEs) directly from data. FDEs, known for their capacity to model non-local dynamics with fewer parameters than integer-order derivatives, can represent complex systems with long-range interactions. Our framework applies deep neural networks as surrogate models for denoising and reconstructing sparse and noisy observations while using Gaussian-Jacobi quadrature to handle the challenges posed by singularities in fractional derivatives. To optimize both the sparse coefficients and fractional order, we employ an alternating optimization approach that combines sparse regression with global optimization techniques. We validate the framework across various datasets, including synthetic anomalous diffusion data, experimental data on the creep behavior of frozen soils, and single-particle trajectories modeled by L\'{e}vy motion. Results demonstrate the framework's robustness in identifying the structure of FDEs across diverse noise levels and its capacity to capture integer-order dynamics, offering a flexible approach for modeling memory effects in complex systems.