Gradient Descent Provably Optimizes Over-parameterized Neural Networks.
Simon S. DuXiyu ZhaiBarnabás PóczosAarti SinghPublished in: CoRR (2018)
Keyphrases
- neural network
- pattern recognition
- cost function
- learning rules
- artificial neural networks
- objective function
- genetic algorithm
- decision trees
- neural nets
- worst case
- back propagation
- neural network model
- loss function
- multi layer
- multilayer perceptron
- feed forward
- fault diagnosis
- fuzzy logic
- knowledge base
- self organizing maps
- associative memory
- lower bound
- biologically inspired
- case study
- fuzzy systems
- neural learning