In many real-world classification problems the input contains a large number of potentially irrelevant features. This paper proposes a new Bayesian framework for determining the r...
Yuan (Alan) Qi, Thomas P. Minka, Rosalind W. Picar...
Learning the user’s semantics for CBIR involves two different sources of information: the similarity relations entailed by the content-based features, and the relevance relatio...
Abstract. Many parallel scienti c applications have dynamic and irregular computational structure. However, most such applications exhibit persistence of computational load and com...
Milind A. Bhandarkar, Robert Brunner, Laxmikant V....
Background: The prediction of protein-protein binding site can provide structural annotation to the protein interaction data from proteomics studies. This is very important for th...
While conventional malware detection approaches increasingly fail, modern heuristic strategies often perform dynamically, which is not possible in many applications due to related ...