AdaBoost is a well known, effective technique for increasing the accuracy of learning algorithms. However, it has the potential to overfit the training set because its objective i...
We present the Distributed Executor for RSES (DIXER) which is a supplementary software for the Rough Set Exploration System (RSES). It takes an advantage of grid computing paradigm...
When dealing with information overload from the Internet, such as the classification of Web pages and the filtering of email spam, a new technique called cotraining has been shown...
Co-training is a semi-supervised technique that allows classifiers to learn with fewer labelled documents by taking advantage of the more abundant unclassified documents. However, ...
A variational level set method is developed for the supervised classification problem. Nonlinear classifier decision boundaries are obtained by minimizing an energy functional tha...