Efficient estimation of dynamically optimal learning rate using higher order derivatives.
Xiao-Hu YuPublished in: ICNN (1996)
Keyphrases
- higher order
- learning rate
- learning algorithm
- multilayer neural networks
- convergence rate
- pairwise
- rapid convergence
- convergence speed
- data mining
- error function
- genetic algorithm
- activation function
- hidden layer
- markov random field
- optimal solution
- estimation error
- particle swarm optimization
- evolutionary algorithm
- middle layer
- adaptive learning rate
- delta bar delta