Partially-observable Markov decision processes (POMDPs) provide a powerful model for sequential decision-making problems with partially-observed state and are known to have (appro...
Unsolicited commercial or bulk emails or emails containing viruses pose a great threat to the utility of email communications. A recent solution for filtering is reputation systems...
Yuchun Tang, Sven Krasser, Dmitri Alperovitch, Pau...
In the context of large databases, data preparation takes a greater importance : instances and explanatory attributes have to be carefully selected. In supervised learning, instanc...
Background: The receiver operating characteristic (ROC) curve is a fundamental tool to assess the discriminant performance for not only a single marker but also a score function c...
Classification is one of the most essential tasks in data mining. Unlike other methods, associative classification tries to find all the frequent patterns existing in the input...