Understanding the training of PINNs for unsteady flow past a plunging foil through the lens of input subdomain level loss function gradients.
Rahul SundarDidier LucorSunetra SarkarPublished in: CoRR (2024)
Keyphrases
- loss function
- stochastic gradient descent
- flow field
- pairwise
- support vector
- learning to rank
- logistic regression
- reproducing kernel hilbert space
- convex loss functions
- solution path
- risk minimization
- training examples
- training samples
- supervised learning
- optical flow
- empirical risk
- input data
- training process
- semi supervised
- training set
- boosting framework
- feature space