Data-driven discovery of self-similarity using neural networks

Abstract

Finding self-similarity is a key step for understanding the governing law behind complex physical phenomena. Traditional methods for identifying self-similarity often rely on specific models, which can introduce significant bias. In this paper, we present a novel neural network-based approach that discovers self-similarity directly from observed data, without presupposing any models. The presence of self-similar solutions in a physical problem signals that the governing law contains a function whose arguments are given by power-law monomials of physical parameters, which are characterized by power-law exponents. The basic idea is to enforce such particular forms structurally in a neural network in a parametrized way. We train the neural network model using the observed data, and when the training is successful, we can extract the power exponents that characterize scale-transformation symmetries of the physical problem. We demonstrate the effectiveness of our method with both synthetic and experimental data, validating its potential as a robust, model-independent tool for exploring self-similarity in complex systems.

21 pages, 18 figures, 5 tables

References in corpus (11)

• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• A Measure of data-collapse for scaling
• Detection of phase transition via convolutional neural network
• Scaling and data collapse from local moments in frustrated disordered quantum spin systems
• Scaling properties of granular rheology near the jamming transition
• Symmetry meets AI
• SymmetryGAN: Symmetry Discovery with Deep Learning
• Scaling Function, Universality and Analytical Solutions of Generalized One-Species Population Dynamics Models
• Relaxation and Intermediate Asymptotics of a Rectangular Trench in a Viscous Film
• Critical scaling of diffusion coefficients and size of rigid clusters of soft athermal particles under shear
• A framework for crossover of scaling law as a self-similar solution : dynamical impact of viscoelastic board