Semi-Supervised Active Learning for Support Vector Machines: A Novel Approach that Exploits Structure Information in Data.
Tobias ReitmaierAdrian CalmaBernhard SickPublished in: CoRR (2016)
Keyphrases
- semi supervised
- raw data
- active learning
- support vector
- background knowledge
- labeled data
- historical data
- complex data
- information sources
- structured information
- unlabeled data
- collected data
- complex structures
- structural information
- prior knowledge
- training data
- data analysis
- sensitive information
- sensor data
- data sets
- semi supervised learning
- computer systems
- end users
- training samples
- data structure
- information resources
- spatial data
- learning algorithm
- knowledge discovery
- web data
- temporal information
- global information
- data collection
- high dimensional data
- log data
- machine learning
- feature selection
- network structure
- multiple sources
- pairwise
- multimedia data
- domain experts
- domain knowledge
- supervised learning
- data points