In the context of the rapidly evolving landscape of vehicular networks, it is highly important to ensure robust security measures in order to protect against cyber threats and anomalies. This paper presents a review of machine learning (ML)-based intrusion detection systems (IDS) designed specifically for vehicular networks. The ML methodologies employed, including supervised, unsupervised, and hybrid approaches, are analysed and compared, with a particular focus on their efficacy in detecting a wide range of intrusion types. The performance of these systems is evaluated by in-depth analysis of key metrics such as detection accuracy, false positive rates, and computational efficiency. Furthermore, the integration of ML-based IDS with existing vehicular communication protocols is explored to enhance overall network resilience. The findings indicate that advanced ML techniques, such as deep learning and ensemble methods, significantly enhance the detection capabilities and adaptability of IDS in dynamic vehicular environments. The objective of this review is to provide researchers and practitioners with valuable insights and guidance on the development of more effective and scalable anomaly detection solutions for secure vehicular networks.