Estimating the Number of Hidden Neurons in a Feedforward Network Using the Singular Value Decomposition.
Eu Jin TeohCheng XiangKay Chen TanPublished in: ISNN (1) (2006)
Keyphrases
- singular value decomposition
- feed forward
- hidden layer
- activation function
- singular values
- back propagation
- neural nets
- artificial neural networks
- hidden neurons
- least squares
- output layer
- dimension reduction
- neural network
- recurrent neural networks
- dimensionality reduction
- principal component analysis
- network architecture
- pattern recognition
- input output
- natural images
- support vector
- feature extraction
- feature selection
- computer vision