Intrusion Detection System (IDS) has become necessary for the security and privacy of a system and it takes a major role in network security because of its detection capacity to various types of attacks in the network domain. Recently, Support Vector Machines (SVM) has been applied to provide useful solutions for intrusion detection systems. With its many variants for classification, SVM is a state-of-theart machine learning algorithm and its performance depends on selection of the appropriate parameters. In this paper, we propose a model based on linear and nonlinear kernel SVMs using NSL-KDD dataset. The parameters for SVM are described in the tabular manner. Then by using the NSL-KDD dataset, our model gives the best result i.e., 100% for accuracy (Both Quadratic and Cubic SVMs).