A Parallel Implementation of a Growing SOM Promoting Independent Neural Networks over Distributed Input Space.
John HammondDaniel MacLeanIren ValovaPublished in: IJCNN (2006)
Keyphrases
- input space
- parallel implementation
- neural network
- feature space
- dimensionality reduction
- input data
- kernel function
- hyperplane
- low dimensional
- high dimensional
- input patterns
- self organizing maps
- parallel computers
- training set
- parallel computation
- high dimensional feature space
- high dimensional data
- distributed systems
- pattern recognition
- data points
- linearly separable
- parallel implementations
- class labels
- k nearest neighbor
- back propagation
- knn
- receptive fields
- multi layer perceptron
- data mining
- recurrent neural networks
- output space
- nearest neighbor
- distributed memory
- principal component analysis
- database