Matrix Smoothing: A Regularization For Dnn With Transition Matrix Under Noisy Labels.
Xianbin LvDongxian WuShu-Tao XiaPublished in: ICME (2020)
Keyphrases
- transition matrix
- markov chain
- smoothing parameter
- directed graph
- training data
- generalized cross validation
- robust statistics
- regularization parameter
- markov decision process
- label noise
- transition probabilities
- machine learning
- total variation regularization
- finite state
- markov model
- bilateral filtering
- pairwise
- training set
- class labels
- state space
- rank minimization
- multi label
- dynamic programming