Bias-Variance Trade-off in Physics-Informed Neural Networks with Randomized Smoothing for High-Dimensional PDEs.
Zheyuan HuZhouhao YangYezhen WangGeorge Em KarniadakisKenji KawaguchiPublished in: CoRR (2023)
Keyphrases
- bias variance
- trade off
- high dimensional
- neural network
- bias variance decomposition
- diffusion equation
- partial differential equations
- pattern recognition
- similarity search
- nearest neighbor
- low dimensional
- level set
- computer science
- nonlinear diffusion
- anisotropic diffusion
- bias variance analysis
- image denoising
- dimensionality reduction
- data points
- numerical solution
- artificial neural networks
- mumford shah model
- feature space
- image processing
- k nearest neighbor
- back propagation
- support vector machine
- knn
- training process
- diffusion process
- probabilistic model
- low variance
- multiscale