Training a feed-forward network with incomplete data due to missing input variables.
Roelof K. BrouwerWitold PedryczPublished in: Appl. Soft Comput. (2003)
Keyphrases
- incomplete data
- feed forward
- input variables
- artificial neural networks
- recurrent networks
- output layer
- missing data
- back propagation
- hidden neurons
- missing values
- feed forward neural networks
- hidden layer
- recurrent neural networks
- learning bayesian networks
- neural network
- bayesian networks
- neural nets
- neural network model
- em algorithm
- variable selection
- spiking neural networks
- feedforward neural networks
- training algorithm
- activation function
- spiking neurons
- fuzzy rules
- multi layer perceptron
- membership functions
- genetic algorithm
- multilayer perceptron
- number of input variables
- incomplete data sets
- expert systems
- learning algorithm
- computational intelligence
- machine learning
- data sets