Symbolic Regression for PDEs using Pruned Differentiable Programs
This addresses the interpretability limitation of black-box neural PDE solvers for researchers and engineers in physics and engineering, though it is incremental as it builds on existing PINN and symbolic regression methods.
The paper tackles the problem of generating interpretable analytical expressions for solutions of partial differential equations (PDEs) by combining physics-informed neural networks (PINNs) with symbolic regression, achieving a 95.3% average reduction in parameters and 7.81% average accuracy improvement while matching PINN performance on complex PDEs like Navier-Stokes flows.
Physics-informed Neural Networks (PINNs) have been widely used to obtain accurate neural surrogates for a system of Partial Differential Equations (PDE). One of the major limitations of PINNs is that the neural solutions are challenging to interpret, and are often treated as black-box solvers. While Symbolic Regression (SR) has been studied extensively, very few works exist which generate analytical expressions to directly perform SR for a system of PDEs. In this work, we introduce an end-to-end framework for obtaining mathematical expressions for solutions of PDEs. We use a trained PINN to generate a dataset, upon which we perform SR. We use a Differentiable Program Architecture (DPA) defined using context-free grammar to describe the space of symbolic expressions. We improve the interpretability by pruning the DPA in a depth-first manner using the magnitude of weights as our heuristic. On average, we observe a 95.3% reduction in parameters of DPA while maintaining accuracy at par with PINNs. Furthermore, on an average, pruning improves the accuracy of DPA by 7.81% . We demonstrate our framework outperforms the existing state-of-the-art SR solvers on systems of complex PDEs like Navier-Stokes: Kovasznay flow and Taylor-Green Vortex flow. Furthermore, we produce analytical expressions for a complex industrial use-case of an Air-Preheater, without suffering from performance loss viz-a-viz PINNs.