Feature selection for time series prediction - A combined filter and wrapper approach for neural networks.
Sven F. CroneNikolaos KourentzesPublished in: Neurocomputing (2010)
Keyphrases
- feature selection
- neural network
- wrapper method
- neural network model
- artificial neural networks
- preprocessing step
- recurrent neural networks
- text categorization
- moving average
- feature ranking
- sequential forward
- text classification
- pattern recognition
- support vector machine
- high dimensionality
- function approximation
- feature selection algorithms
- non stationary
- feature extraction
- feature space
- feature subset
- feature set
- irrelevant features
- classification models
- classification accuracy
- gene expression data
- mutual information
- neuro fuzzy
- information gain
- back propagation
- model selection
- multi class
- feed forward
- prediction model
- knn
- activation function
- support vector
- redundant features
- reinforcement learning
- dimensionality reduction