xdfeng7370 Goto Github PK

deep-ritz-method

Reproduce the first two numerical experiments(Pytorch)

gradient-flow-in-pinns

Depict the gradient information in pinns

multi-fidelity-neural-network-

multiscale_hw

Multiscale homework

pinns

PyTorch Implementation of Physics-informed Neural Networks

pyapprox

pymc

Bayesian Modeling and Probabilistic Programming in Python

shenfun

High performance computational platform in Python for the spectral Galerkin method

some-numerical-experiments-on-drm-dgm-and-wan

I reproduce three simple PDE based on Pytorch , including irregular domain and non-smooth function. In first two examples, I test deep ritz method and PINNs. In the last examples, I test WAN method (weak adversarial networks).

some_tricks_in_deep_learning

I reproduce three numerical experiments. The first example depicts the symmetric structure in neural network; the second example shows the adaptive activation function, namely, $\sigma(x)\to \sigma(ax)$, where $a$ is a learnable parameter. The last example shows the spectral bias in the DNN.

uncertainty-toolbox

Uncertainty Toolbox: a Python toolbox for predictive uncertainty quantification, calibration, metrics, and visualization

xdfeng7370.github.io