Interpretable Solutions for Multi-Physics PDEs Using T-NNGP

Authors: Lulu Cao, Zexin Lin, Kay Chen Tan, Min Jiang

AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on three types of PDEs demonstrate that our method can reliably obtain human-understandable symbolic expressions that fit both the PDEs and the numerical solutions from traditional methods. This work advances multiphysics simulation by enhancing our ability to derive approximate symbolic solutions for PDEs, thereby improving our understanding of complex physical phenomena. In this section, experiments on three types of systems of PDEs are conducted to evaluate the performance of the proposed T-NNGP.
Researcher Affiliation Academia Lulu Cao1,2, 3, Zexin Lin1,2, Kay Chen Tan3, Min Jiang1,2* 1 Key Laboratory of Multimedia Trusted Perception and Efficient Computing, Ministry of Education of China, Xiamen University 2 School of Informatics, Key Laboratory of Digital Protection and Intelligent Processing of Intangible Cultural Heritage of Fujian and Taiwan, Ministry of Culture and Tourism, Xiamen University 3 Department of Data Science and Artificial Intelligence, The Hong Kong Polytechnic University
Pseudocode Yes Algorithm 1: T-NNGP Input: pdes (A system of PDEs to be solved), cu0 (the degree of constraint on each unknown function), f (Fitness function) Output: T (Optimal symbolic expression solutions)
Open Source Code Yes Code https://github.com/grassdeerdeer/T-NNGP
Open Datasets No In this paper, the physical dataset is generated by solving the systems of PDEs utilising the finite element method (FEM) (Permann et al. 2020; Li and Chen 2019).
Dataset Splits Yes For the symbolic regression method (Phy SO and TNNGP), 100 data points are randomly selected to generate expression. For Deep M&Mnet, 10% of the dataset is taken as training data. Compared to the entire dataset, very little data is used for training, so we use all the data in the dataset for evaluation.
Hardware Specification No No specific hardware details (like GPU/CPU models or specific cloud resources) were mentioned in the paper for running experiments.
Software Dependencies No The paper mentions methods like deep reinforcement learning, recursive neural networks (RNN), and genetic programming, but does not provide specific version numbers for any software libraries, frameworks, or languages used.
Experiment Setup No The paper states, "Table 3 in Appendix D lists the main parameters setting of T-NNGP." However, the content of Appendix D, including Table 3, is not provided in the main text of the paper.