Advanced vehicles and cars carry contemporary computing, control, and communication devices and technologies that turn them into intelligent self-drive systems. However, Due to over-dependency on communications and computing technologies, smart cars may become vulnerable to intelligently-designed car hacking cyber-attacks. Machine-learning-based techniques are extensively employed to detect such attacks. This paper proposes a Euclidean distance-based machine learning technique to detect car hacking denial of service (DoS) cyber-attacks.