Dynamical Change of the Perceiving Properties of Neural Networks as Training with Noise and Its Impact on Pattern Recognition.
Roman NemkovPublished in: YSIP (2014)
Keyphrases
- pattern recognition
- neural network
- training algorithm
- training process
- train a neural network
- multi layer perceptron
- backpropagation algorithm
- image processing
- feed forward neural networks
- back propagation
- signal processing
- training phase
- noise model
- pattern recognition problems
- feedforward neural networks
- artificial neural networks
- neural network structure
- gaussian noise
- noise level
- machine learning
- feature extraction
- neural network training
- learning rules
- modular neural networks
- computer vision
- multilayer perceptron
- signal to noise ratio
- missing data
- test set
- genetic algorithm
- dimensionality reduction
- rough sets
- training set
- image analysis
- training examples
- data sets
- training and testing data
- error back propagation
- fuzzy logic
- image noise
- noise reduction
- multi layer
- hidden layer