Automated gastric cancer diagnosis on H&E-stained sections; ltraining a classifier on a large scale with multiple instance machine learning.
Eric CosattoPierre-François LaquerreChristopher MalonHans Peter GrafAkira SaitoTomoharu KiyunaAtsushi MarugameKen'ichi KamijoPublished in: Medical Imaging: Digital Pathology (2013)
Keyphrases
- multiple instance
- gastric cancer
- machine learning
- multiple instance learning
- learning problems
- early detection
- semi supervised learning
- supervised learning
- concept learning
- learning algorithm
- decision trees
- application of data mining
- real valued
- machine learning algorithms
- support vector machine
- feature selection
- training data
- feature space
- learning tasks
- relational learning
- class labels
- inductive logic programming
- machine learning methods
- information extraction
- feature set
- multi class
- multiple kernel learning
- training samples
- text mining
- maximum margin
- svm classifier
- genetic programming
- kernel methods
- natural language processing
- inductive learning
- active learning
- training set
- knowledge discovery
- semi supervised
- natural language
- data analysis
- pairwise
- feature vectors
- loss function
- multi modal