yingtaoluo/pde-discovery-with-evolutionary-tree-search

Early version codes for "Symbolic genetic algorithm for discovering open-form partial differential equations" published in Physical Review Research

PDE-Discovery-with-Neural-Evolutionary-Tree-Search

It is an implementation of "Symbolic genetic algorithm for discovering open-form partial differential equations (SGA-PDE)" by the original author. The official repository is available [here] and the published paper is [here].

This algorithm can efficiently establish mathematical equation with tree representation to fit existing dataset. We can discover partial derivative equation (PDE) with nested forms without any prior knowledge about the equation candidate terms. The genetic algorithm provides exploration and optimization on symbolic mathematics, the forest provides a concise way to represent equations, and machine learning provides a fitting accuracy indicator.

Version: This repo is an early version that contains the basic structure of the algorithm/code. For re-implementing the paper, please see the official repository.

pde-discovery-with-evolutionary-tree-search's Issues

Steps to follow

Hello, I came across your repo while searching for data driven discovery of PDEs and I must say I am really impressed by the work you have done. Is it possible that you can suggest the steps one needs to follow to implement your code on any other dataset?

Also can you share the code for the nonparametric sparse regression that you have mentioned in the README of the repo?