DScGANS: Integrate Domain Knowledge in Training Dual-Path Semi-supervised Conditional Generative Adversarial Networks and S3VM for Ultrasonography Thyroid Nodules Classification.
Wenkai YangJuanjuan ZhaoYan QiangXiaotang YangYunyun DongQianqian DuGuohua ShiMuhammad Bilal ZiaPublished in: MICCAI (4) (2019)
Keyphrases
- semi supervised
- domain knowledge
- supervised learning
- unsupervised learning
- training phase
- training process
- training set
- generative model
- lung nodules
- discriminative learning
- discriminative classifiers
- training samples
- co training
- feature vectors
- text classification
- classification accuracy
- support vector
- decision trees
- active learning
- classification algorithm
- discriminative training
- partially labeled data
- machine learning
- semi supervised classification
- image classification
- feature extraction
- semi supervised learning
- labelled training data
- neural network
- training examples
- class labels
- feature selection
- training data
- multi agent
- feature space
- bayesian networks
- unsupervised clustering
- pairwise constraints
- pairwise
- support vector machine
- medical diagnosis
- medical images
- shortest path
- unlabeled data