Penalized Entropy: a novel loss function for uncertainty estimation and optimization in medical image classification.
Dehua FengXi ChenXiaoyu WangJiahuan LvLing BaiShu ZhangZhiguo ZhouPublished in: CBMS (2022)
Keyphrases
- loss function
- image classification
- risk minimization
- stochastic gradient descent
- pairwise
- loss minimization
- support vector
- image representation
- logistic regression
- learning to rank
- information theoretic
- feature extraction
- reproducing kernel hilbert space
- statistical decision theory
- coordinate descent method
- convex loss functions
- empirical risk
- hinge loss
- mutual information
- bag of words
- least squares
- boosting framework
- image features
- regularization term
- maximum likelihood
- boosting algorithms
- class specific
- optimization problems