Input space bifurcation manifolds of recurrent neural networks.
Robert HaschkeJochen J. SteilPublished in: Neurocomputing (2005)
Keyphrases
- recurrent neural networks
- input space
- low dimensional
- feature space
- high dimensional data
- high dimensional
- low dimensional manifolds
- dimensionality reduction
- manifold learning
- neural network
- kernel function
- recurrent networks
- data points
- embedding space
- reservoir computing
- hidden layer
- input data
- echo state networks
- feed forward
- piecewise linear
- artificial neural networks
- k nearest neighbor
- hyperplane
- euclidean space
- linearly separable
- feedforward neural networks
- nearest neighbor
- class labels
- nonlinear dynamic systems
- manifold structure
- data sets
- nonlinear dimensionality reduction
- knn
- kernel pca
- input patterns
- data mining
- output space
- classification accuracy
- feature set
- geometric structure
- training set
- computer vision
- learning algorithm
- smooth functions
- principal component analysis