A Regularization Post Layer: An Additional Way How to Make Deep Neural Networks Robust.

Jan Vanek Jan Zelinka Daniel Soutner Josef Psutka

Published in: SLSP (2017)

Keyphrases

neural network
multi layer
pattern recognition
rank minimization
associative memory
single layer
multiple layers
fuzzy logic
computationally efficient
prior information
neural network model
feed forward
recurrent neural networks
parameter selection