Early stabilizing feature importance for TensorFlow deep neural networks.
Jeff HeatonSteven McElweeJames B. FraleyJames CannadyPublished in: IJCNN (2017)
Keyphrases
- feature importance
- neural network
- random forest
- pattern recognition
- artificial neural networks
- relevance feedback
- fuzzy logic
- semi supervised
- feature selection
- neural nets
- neural network model
- feed forward
- back propagation
- recurrent neural networks
- multilayer perceptron
- higher order
- nonlinear systems
- self organizing maps
- multi class
- support vector machine
- learning process
- computer vision