Neural Networks with Local Receptive Fields and Superlinear VC Dimension
Michael SchmittPublished in: Electron. Colloquium Comput. Complex. (2001)
Keyphrases
- receptive fields
- vc dimension
- neural network
- hidden layer
- biologically inspired
- sample complexity
- upper bound
- natural images
- lower bound
- image representation
- back propagation
- sample size
- inductive inference
- worst case
- pattern recognition
- artificial neural networks
- multilayer perceptron
- fuzzy logic
- input space
- space time
- multiscale
- compression scheme
- saliency map
- recurrent neural networks
- neural nets
- feed forward
- neural network model
- learning problems
- fault diagnosis
- nearest neighbor
- higher order
- support vector