On transversality of bent hyperplane arrangements and the topological expressiveness of ReLU neural networks.
J. Elisenda GrigsbyKathryn LindseyPublished in: CoRR (2020)
Keyphrases
- hyperplane
- neural network
- support vector
- training samples
- input space
- pattern recognition
- data points
- feature space
- maximal margin
- linear classifiers
- incremental learning algorithm
- decision boundary
- kernel function
- convex hull
- support vector machine
- linear separability
- principal components
- average distance
- classification procedure
- data sets
- back propagation
- support vectors
- artificial neural networks
- linearly separable
- machine learning
- locality sensitive
- nearest neighbor
- weight vector
- normal vectors
- reinforcement learning