Self-scalable Tanh (Stan): Faster Convergence and Better Generalization in Physics-informed Neural Networks.
Raghav GnanasambandamBo ShenJihoon ChungXubo YueZhenyu James KongPublished in: CoRR (2022)
Keyphrases
- faster convergence
- neural network
- convergence speed
- global optimum
- step size
- pattern recognition
- artificial neural networks
- global optimization
- pso algorithm
- fuzzy logic
- back propagation
- adaptive resonance theory
- training algorithm
- fuzzy systems
- recurrent neural networks
- convergence rate
- computer science
- multilayer perceptron
- radial basis function
- multi layer
- training process
- artificial intelligence
- feed forward
- associative memory
- neural nets
- neural network model
- lightweight
- highly scalable
- genetic algorithm
- data mining