A novel training method has been proposed for increasing efficiency and generalization of support vector machine (SVM). The efficiency of SVM in classification is directly determi...
Data sets in large applications are often too massive to fit completely inside the computer's internal memory. The resulting input/output communication (or I/O) between fast ...
This paper explores various aspects of the image decomposition problem using modern variational techniques. We aim at splitting an original image f into two components u and v, whe...
We consider Cooperative Intrusion Detection System (CIDS) which is a distributed AIS-based (Artificial Immune System) IDS where nodes collaborate over a peer-to-peer overlay netwo...
This paper presents a new deterministic approximation technique in Bayesian networks. This method, "Expectation Propagation," unifies two previous techniques: assumed-de...