Width is Less Important than Depth in ReLU Neural Networks.
Gal VardiGilad YehudaiOhad ShamirPublished in: COLT (2022)
Keyphrases
- neural network
- pattern recognition
- neural network model
- artificial neural networks
- back propagation
- fuzzy logic
- depth map
- fuzzy systems
- depth information
- self organizing maps
- case study
- feed forward
- artificial intelligence
- multi view
- data sets
- computer vision
- multilayer perceptron
- genetic algorithm
- depth estimation
- feedforward neural networks
- competitive learning
- neural network is trained