Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topology.
Bastian RieckMatteo TogninalliChristian BockMichael MoorMax HornThomas GumbschKarsten M. BorgwardtPublished in: ICLR (Poster) (2019)
Keyphrases
- neural network
- network architecture
- artificial neural networks
- computational complexity
- complexity measures
- neural computation
- pattern recognition
- connectionist models
- competitive learning
- recurrent neural networks
- neural model
- learning rules
- worst case
- similarity measure
- fault diagnosis
- neural learning
- hebbian learning
- back propagation
- distance measure
- fuzzy logic
- topology preserving
- deep learning
- computational intelligence
- distributed representations
- neural fuzzy
- computational cost
- self organizing maps
- fuzzy neural network
- training algorithm
- multi layer
- correlation coefficient
- radial basis function
- decision problems