NeuRecover: Regression-Controlled Repair of Deep Neural Networks with Training History.
Shogo TokuiSusumu TokumotoAkihito YoshiiFuyuki IshikawaTakao NakagawaKazuki MunakataShinji KikuchiPublished in: CoRR (2022)
Keyphrases
- neural network
- training process
- training algorithm
- multi layer perceptron
- feed forward neural networks
- regression model
- feedforward neural networks
- neural network training
- training phase
- radial basis function network
- pattern recognition
- learning machines
- regression problems
- linear regression
- test set
- feature selection
- back propagation
- model selection
- recurrent networks
- backpropagation algorithm
- locally weighted
- training set
- neural network structure
- deep learning
- training patterns
- deep architectures
- error back propagation
- supervised learning
- fuzzy logic
- artificial neural networks
- gaussian processes
- support vector regression
- neural nets
- recurrent neural networks
- neural network model
- training samples
- online learning
- damage assessment
- support vector
- training data
- machine learning
- data sets