Matrix Smoothing: A Regularization for DNN with Transition Matrix under Noisy Labels.
Xianbin LvDongxian WuShu-Tao XiaPublished in: CoRR (2020)
Keyphrases
- transition matrix
- markov chain
- directed graph
- smoothing parameter
- bilateral filtering
- training data
- markov decision process
- label noise
- state space
- robust statistics
- finite state
- markov model
- generalized cross validation
- transition probabilities
- class labels
- training process
- missing data
- pairwise
- trace norm
- total variation regularization
- multi agent
- reinforcement learning
- bayesian networks