Interval Weight-Based Abstraction for Neural Network Verification.
Fateh BoudardaraAbderraouf BoussifPierre-Jean MeyerMohamed GhazelPublished in: SAFECOMP Workshops (2022)
Keyphrases
- neural network
- weight update
- hardware designs
- back propagation
- artificial neural networks
- neural network model
- interval data
- model checking
- neural network is trained
- pattern recognition
- global exponential stability
- formal verification
- data sets
- bounded model checking
- self organizing maps
- backpropagation neural network
- asynchronous circuits
- multi layer
- network model
- face verification
- multi layer perceptron
- weight assignment
- feedforward neural networks
- network architecture
- genetic algorithm
- training algorithm
- fuzzy logic
- neural nets
- verification method
- type fuzzy logic systems
- functional verification
- high level
- fault diagnosis
- feed forward
- bp neural network
- weighting scheme
- hidden layer
- data abstraction
- learning vector quantization
- neural model
- activation function