Accelerating Hessian-free optimization for Deep Neural Networks by implicit preconditioning and sampling.
Tara N. SainathLior HoreshBrian KingsburyAleksandr Y. AravkinBhuvana RamabhadranPublished in: ASRU (2013)
Keyphrases
- neural network
- iterative methods
- pattern recognition
- artificial neural networks
- fuzzy logic
- genetic algorithm
- edge preserving
- optimization method
- optimization algorithm
- fault diagnosis
- constrained optimization
- optimization problems
- highly non linear
- back propagation
- sample size
- linear programming
- optimization methods
- random sampling
- multi layer
- fuzzy systems
- sampling algorithm
- optical flow computation
- learning algorithm
- hessian matrix