Mitigating Neural Network Overconfidence with Logit Normalization.
Hongxin WeiRenchunzi XieHao ChengLei FengBo AnYixuan LiPublished in: ICML (2022)
Keyphrases
- neural network
- artificial neural networks
- pattern recognition
- back propagation
- network model
- normalization method
- self organizing maps
- fuzzy neural network
- regression analysis
- training algorithm
- bp neural network
- neural network is trained
- information systems
- genetic algorithm
- machine learning
- image reconstruction from projections
- back propagation neural network
- feed forward neural networks
- activation function
- hybrid neural network
- hidden layer
- multi layer
- neural nets
- neural network model
- fault diagnosis
- preprocessing
- computer vision