ENSEMBLE LEARNING FOR MULTIPLE DATA MINING PROBLEMS
MetadataShow full item record
Data mining practitioners often face problems of the unavailability of all training data at the same time and the inability to process a large amount of data due to constraints such as lack of adequate system memory. Building a data mining system with whatever data available at a certain time is data is a practical solution. Our hypothesis is that a learning model should be able to update on incoming data in an incremental manner. Another challenge arises when new classes are introduced into a trained system during testing because the learned model does not have an ability to handle unknown ...