Knowledge Transfer via Distillation of Activation Boundaries Formed by Hidden Neurons.
Byeongho HeoMinsik LeeSangdoo YunJin Young ChoiPublished in: CoRR (2018)
Keyphrases
- knowledge transfer
- hidden neurons
- generalization ability
- artificial neural networks
- activation function
- feedforward neural networks
- neural network
- input variables
- knowledge sharing
- transfer learning
- training speed
- hidden layer
- neural network model
- learning algorithm
- learning tasks
- feed forward
- data sets
- recurrent neural networks
- support vector
- learning rate
- feature extraction
- data mining