FAT: Training Neural Networks for Reliable Inference Under Hardware Faults.
Ussama ZahidGiulio GambardellaNicholas J. FraserMichaela BlottKees A. VissersPublished in: ITC (2020)
Keyphrases
- neural network
- training process
- fault diagnosis
- training algorithm
- feed forward neural networks
- pattern recognition
- feedforward neural networks
- backpropagation algorithm
- hardware and software
- low cost
- real time
- structured prediction
- multi layer perceptron
- training set
- back propagation
- neural network structure
- radial basis function network
- embedded systems
- probabilistic inference
- fault detection
- cost effective
- feed forward
- artificial neural networks
- bayesian networks
- neural network training
- multiple faults
- hardware architecture
- training patterns
- training phase
- multi layer
- fuzzy systems
- genetic algorithm
- neural nets
- recurrent neural networks
- fuzzy logic
- test set
- data sets
- fault model
- image processing
- recurrent networks
- online learning
- computer software
- root cause
- self organizing maps
- multilayer perceptron
- associative memory
- bayesian inference
- hardware implementation
- model based diagnosis