Comparison of non-linear activation functions for deep neural networks on MNIST classification task.
Dabal PedamontiPublished in: CoRR (2018)
Keyphrases
- activation function
- neural network
- pattern recognition
- feed forward
- back propagation
- feed forward neural networks
- classification accuracy
- neural nets
- neural architecture
- multi layer perceptron
- hidden layer
- multilayer perceptron
- artificial neural networks
- feedforward neural networks
- training phase
- feature space
- text classification
- feature extraction
- learning rate
- handwritten digits
- machine learning
- basis functions
- decision trees
- feature selection
- class labels
- supervised learning
- feature vectors
- training set
- connection weights
- neural network model
- classification algorithm
- machine learning algorithms
- particle swarm optimization
- fuzzy logic
- decision making
- hidden nodes
- genetic algorithm