At present, use of the internet has increased in our modern technological life. Thus a lot of sensitive and critical data
is being transmitted which increases the number of intrusion attempt over the network system. So cyber security has
become a vital issue to prevent threats. Intrusion Detection System (IDS) is a dynamic security system to recognize
suspicious and malicious activities or policy violations. Attackers try to bypass the security firewall, therefore, the
need of IDS is increasing. IDS is used to detect the misuses in the system. Various machine learning techniques have
been used to detect intrusion within the systems. Machine Learning focuses on classification and prediction based on
known properties of dataset. This study has focused on feature selection and classification model for intrusion
detection based on machine learning techniques. The NSL-KDD dataset has been used for experiment and WEKA
(Waikato environment for knowledge analysis) as experimental tool. For performance measures of the classification
models Precision, Recall and Receiver Operating Characteristics (ROC) curve evaluation matrices have been used.