In this letter, we consider the binary classification problem via distributed Support Vector Machines (SVMs), where idea is to train a network of agents, with limited share data, cooperatively learn SVM classifier for global database. Agents only processed information regarding parameters and gradient local loss functions instead their raw data. contrast existing work, propose continuous-time a...