Orthogonalising gradients to speed up neural network optimisation.
Mark TuddenhamAdam Prügel-BennettJonathan HarePublished in: CoRR (2022)
Keyphrases
- image segmentation
- neural network
- genetic algorithm
- back propagation
- gradient vector
- neural network model
- artificial neural networks
- fault diagnosis
- feed forward
- neural network is trained
- multi layer
- neural nets
- pattern recognition
- fuzzy logic
- machine learning
- self organizing maps
- multilayer perceptron
- model predictive control
- associative memory
- network model
- global optimisation
- information retrieval
- learning vector quantization
- feed forward neural networks
- learning rules
- hidden layer
- artificial intelligence
- decision making
- prediction model
- recurrent neural networks
- data sets