Regularizing Neural Networks by Penalizing Confident Output Distributions.
Gabriel PereyraGeorge TuckerJan ChorowskiLukasz KaiserGeoffrey E. HintonPublished in: ICLR (Workshop) (2017)
Keyphrases
- neural network
- desired output
- pattern recognition
- artificial neural networks
- probability distribution
- neural network model
- recurrent neural networks
- machine learning
- decision making
- fuzzy logic
- rbf network
- data sets
- hidden units
- power law
- neural nets
- multilayer perceptron
- feed forward
- random variables
- computational intelligence
- genetic algorithm