MaxDropout: Deep Neural Network Regularization Based on Maximum Output Values.
Claudio Filipi Goncalves dos SantosDanilo ColomboMateus RoderJoão Paulo PapaPublished in: ICPR (2020)
Keyphrases
- neural network
- desired output
- input variables
- artificial neural networks
- back propagation
- neural network model
- probability values
- sigmoid function
- smoothing parameter
- feed forward neural networks
- neural nets
- pattern recognition
- rbf network
- learning vector quantization
- prior information
- feed forward
- recurrent neural networks
- bp neural network
- fuzzy logic
- numerical values
- data sets
- training data
- deep learning
- network architecture
- network model
- multi layer
- self organizing maps