Advanced control systems are becoming more and more sophisticated every day. For safety-critical systems such as chemical processes, nuclear reactors, aircraft and spacecraft, the issue of reliability, acceptable performance, and environmental protection are of particular importance. If a fault occurs, the damages to financial, human, and environmental could be severe. As a result, there is a growing need for online monitoring and Fault Detection, or to identify system faults to improve reliability. Accordingly, it is viable to apply the preliminary signs and to prevent the system from stopping and the occurrence of a catastrophe to a large extent. Fault diagnostic methods can be divided into two general categories of model-based and data-based methods. In this chapter, we will look at the second category, which includes fuzzy logic, neural network, and support vector machines.