A comparative study of Machine Learning algorithms for the detection of vulnerable Python libraries

Detecting the existence of vulnerabilities within source code is an important step in improving the overall security of an organisation and reducing the possibility of an attacker breaching the IT system. This has led to the creation of different vulnerability detection tools and, therefore, to devoting efforts to the study of detection techniques to provide the best results. One of the techniques used for this purpose is those that use Machine Learning and Data Mining models, this being a booming field. Under this premise, this paper presents a comparison of the results obtained with Machine Learning models capable of classifying the vulnerability or non-vulnerability of a real-world source code in Python language.