Seq-SVF: An unsupervised data-driven method for automatically identifying hidden governing equations.
Zhetong WuHongfei YeHongwu ZhangYonggang ZhengPublished in: Comput. Phys. Commun. (2023)
Keyphrases
- data driven
- detection method
- preprocessing
- mathematical model
- high accuracy
- significant improvement
- computational cost
- clustering method
- high precision
- synthetic data
- probabilistic model
- computationally efficient
- unsupervised learning
- classification accuracy
- experimental evaluation
- objective function
- unsupervised manner
- feature set
- classification method
- algebraic equations
- support vector machine svm
- mutual information
- neural network
- cost function
- pairwise
- multiscale
- similarity measure