Upper Bounds on the Approximation Rates of Real-valued Boolean Functions by Neural Networks.
Katerina HlaváckováVera KurkováPetr SavickýPublished in: ICANNGA (1997)
Keyphrases
- real valued
- boolean functions
- upper bound
- neural network
- linear functions
- continuous functions
- lower bound
- uniform distribution
- complex valued
- upper and lower bounds
- worst case
- integer valued
- distributed representations
- dnf formulae
- real valued data
- functional properties
- lower and upper bounds
- recurrent neural networks
- generalization error
- membership queries
- fuzzy logic
- dnf formulas
- sample complexity
- sample size
- vc dimension
- statistical queries
- concept class
- multiple instance
- multi valued