Open set task augmentation facilitates generalization of deep neural networks trained on small data sets.
Wadhah Zai El AmriFelix ReinhartWolfram SchenckPublished in: Neural Comput. Appl. (2022)
Keyphrases
- small data sets
- neural network
- multilayer perceptron
- trained neural networks
- data sets
- training process
- multi layer perceptron
- trained neural network
- auto associative
- backpropagation algorithm
- artificial neural networks
- back propagation
- pattern recognition
- genetic algorithm
- competitive learning
- rule extraction
- multilayer neural network
- deep learning
- activation function
- self organizing maps
- fuzzy logic
- training set
- computer vision
- recurrent neural networks
- neural nets
- neural network model
- radial basis function
- feedforward neural networks
- hidden units
- rough sets
- real time
- data mining
- feature selection
- neural learning
- decision trees
- single layer
- feed forward
- feed forward neural networks
- fault diagnosis
- network architecture
- machine learning