Angle based dynamic learning rate for gradient descent.
Neel MishraPawan KumarPublished in: CoRR (2023)
Keyphrases
- learning rate
- error function
- convergence rate
- natural gradient
- learning algorithm
- update rule
- convergence speed
- cost function
- hidden layer
- multilayer neural networks
- control parameters
- search space
- genetic programming
- objective function
- weight vector
- adaptive learning rate
- step size
- delta bar delta
- training algorithm
- linear combination
- linear programming
- support vector
- feature selection