DeepParticle: Learning invariant measure by a deep neural network minimizing Wasserstein distance on data generated from an interacting particle method.
Zhongjian WangJack XinZhiwen ZhangPublished in: J. Comput. Phys. (2022)
Keyphrases
- prior knowledge
- synthetic data
- neural network
- data sets
- input data
- data analysis
- test data
- similarity measure
- learned models
- neural nets
- correlation analysis
- bayesian methods
- missing data
- training data
- noisy data
- dissimilarity measure
- unsupervised learning
- training process
- background knowledge
- learning algorithm
- data structure
- database
- correlation coefficient
- distance matrix
- missing values
- training samples
- learning process
- statistical methods
- clustering method
- prediction model
- model selection
- distance measure
- k means
- moment invariants
- reinforcement learning
- machine learning