Data-driven prediction of soliton solutions of the higher-order NLSE via the strongly-constrained PINN method.
Yin FangYue-Yue WangWei LiuChao-Qing DaiPublished in: Comput. Math. Appl. (2022)
Keyphrases
- data driven
- higher order
- high accuracy
- objective function
- high order
- high precision
- pairwise
- synthetic data
- significant improvement
- computational cost
- input data
- prediction algorithm
- data sets
- training set
- markov random field
- experimental evaluation
- cost function
- computationally efficient
- segmentation algorithm
- clustering method
- prior knowledge
- missing data
- segmentation method
- computational complexity
- optimization method
- prediction model
- predictive model
- decision trees