The machine learning-based IDS models have the ability to learn the attack patterns from the pre-defined training data automatically and detect malicious activities in the VANET system. The machine learning models also improvise the performance of data-rich VANET environment. Some of the familiar machine learning algorithms are k-means clustering, support vector machine, k-nearest neighbor, and naive Bayes, which are frequently used for intrusion detection for VANETs. The machine learning IDS models are self-adaptive that comprise diverse statistical models for prediction and detection decision-making based on abundant VANET data. The machine learning-based IDSs is classified into three major categories that are supervised, unsupervised, and reinforcement.