Correlated Parameters to Accurately Measure Uncertainty in Deep Neural Networks.
Konstantin PoschJürgen PilzPublished in: CoRR (2019)
Keyphrases
- neural network
- highly correlated
- pattern recognition
- similarity measure
- decision making
- parameter values
- fuzzy logic
- probability measures
- probability measure
- artificial neural networks
- maximum likelihood
- expectation maximization
- back propagation
- nonlinear functions
- neural network model
- uncertain data
- expected utility
- input parameters
- multi layer perceptron
- highly non linear
- network architecture
- multi layer
- neural nets
- recurrent neural networks
- sensitivity analysis
- membership functions
- input data
- support vector machine