Deep Belief Networks (DBN's) are generative models that contain many layers of hidden variables. Efficient greedy algorithms for learning and approximate inference have allow...
An extension of the AdaBoost learning algorithm is proposed and brought to bear on the face detection problem. In each weak classifier selection cycle, the novel totally correctiv...
All-to-all personalized exchange is one of the most dense collective communication patterns and occurs in many important parallel computing/networking applications. In this paper,...
In the area of privacy-preserving data mining, a differentially private mechanism intuitively encourages people to share their data truthfully because they are at little risk of ...
—Counting Bloom Filters (CBFs) are widely used in networking device algorithms. They implement fast set representations to support membership queries with limited error, and supp...