Art: Abstraction Refinement-Guided Training for Provably Correct Neural Networks.
Xuankang LinHe ZhuRoopsha SamantaSuresh JagannathanPublished in: FMCAD (2020)
Keyphrases
- provably correct
- neural network
- training algorithm
- training process
- feedforward neural networks
- multi layer perceptron
- high level
- pattern recognition
- feed forward neural networks
- situation calculus
- error back propagation
- adaptive resonance theory
- artificial neural networks
- multi layer
- neural network training
- feed forward
- software engineering
- fuzzy logic
- training phase
- training set
- radial basis function
- training samples
- support vector machine
- data model
- expert systems
- error estimates
- reinforcement learning