How Shrinking Gradient Noise Helps the Performance of Neural Networks.
Zhun DengJiaoyang HuangKenji KawaguchiPublished in: IEEE BigData (2021)
Keyphrases
- neural network
- pattern recognition
- edge detection
- noise reduction
- gradient estimation
- noisy data
- noise level
- noise model
- fuzzy logic
- signal to noise ratio
- neural nets
- random noise
- noise free
- gaussian noise
- missing data
- back propagation
- artificial neural networks
- multi layer
- gradient information
- median filter
- activation function
- additive noise
- genetic algorithm
- multilayer perceptron
- step edges
- image processing
- training data
- noise removal
- training algorithm
- training process
- neural network model