Predictive models developed by applying Data Mining techniques are used to improve forecasting accuracy in the airline business. In order to maximize the revenue on a flight, the ...
Web page classification is important to many tasks in information retrieval and web mining. However, applying traditional textual classifiers on web data often produces unsatisfyi...
A hypergraph is a generalization of the traditional graph in which the edges are arbitrary non-empty subsets of the vertex set. It has been applied successfully to capture highord...
Traditional classification methods assume that the training and the test data arise from the same underlying distribution. However, in several adversarial settings, the test set is...
Various semi-supervised learning methods have been proposed recently to solve the long-standing shortage problem of manually labeled data in sentiment classification. However, mos...