Semi-supervised active learning for support vector machines: A novel approach that exploits structure information in data.
Adrian CalmaTobias ReitmaierBernhard SickPublished in: Inf. Sci. (2018)
Keyphrases
- raw data
- semi supervised
- active learning
- support vector
- collected data
- structural information
- information sources
- data sets
- historical data
- sensor data
- labeled data
- global information
- data analysis
- data structure
- end users
- data collection
- complex data
- background knowledge
- complex structures
- prior knowledge
- computer systems
- multiple sources
- web data
- training data
- structured information
- data sources
- sensitive information
- information resources
- temporal information
- training examples
- database
- support vector machine
- knowledge discovery
- multimedia data
- unlabeled data
- high dimensional
- feature selection
- data mining techniques
- learning algorithm
- machine learning
- privacy concerns
- multiple types