DALSA: Domain Adaptation for Supervised Learning From Sparsely Annotated MR Images.
Michael GötzChristian WeberFranciszek BinczykJoanna PolanskaRafal TarnawskiBarbara Bobek-BillewiczUllrich KötheJens KleesiekBram StieltjesKlaus H. Maier-HeinPublished in: IEEE Trans. Medical Imaging (2016)
Keyphrases
- mr images
- domain adaptation
- supervised learning
- labeled data
- semi supervised
- semi supervised learning
- unlabeled data
- magnetic resonance
- medical images
- co training
- magnetic resonance images
- training data
- unsupervised learning
- transfer learning
- target domain
- image data
- active learning
- training set
- cross domain
- training examples
- class labels
- phantom images
- brain tumors
- test data
- mr imaging
- sparse representation
- training samples
- learning algorithm
- intensity inhomogeneity
- learning tasks
- text categorization
- document classification
- multiple sources
- test set
- sentiment classification
- reinforcement learning
- data sets
- multi view
- support vector machine
- prior knowledge
- high dimensional