Author(s): Barros, Sara Antunes
Date: 2025
Persistent ID: http://hdl.handle.net/10362/190649
Origin: Repositório Institucional da UNL
Subject(s): Artificial Intelligence; Cybersecurity; Machine Learning; Predictive Modelling; SDG 9 - Industry, innovation and infrastructure; Domínio/Área Científica::Ciências Naturais::Ciências da Computação e da Informação
Dissertation presented as the partial requirement for obtaining a Master's degree in Information Management, specialization in Business Intelligence
This research explores how Artificial Intelligence models can predict malicious network traffic. This is relevant with the increasing number of cyberattacks, considering Artificial Intelligence technology has the ability to protect against them. To do so, it is important to first determine which models have are able to play this defensive role. The focus and objective of this research is to understand the practical use and the predictive power of these models and, with the power of Python, an experiment is conducted to assess whether there is a model that is able to predict malicious network according to several metrics such as accuracy, support, recall, and F1 score considering a public dataset found on Kaggle. The research proves that Artificial Intelligence, and especially Machine Learning models, have the potential to help organizations stay cybersafe against attacks, provided the models are used by professionals who are able to understand not only the models, but also the business in which they are inserted.
