Use this link to cite:
https://hdl.handle.net/2183/46137 Detección e clasificación de ciberataques mediante técnicas de aprendizaxe automática
Loading...
Identifiers
Publication date
Authors
Cela Riveiro, Aarón
Advisors
Other responsabilities
Universidade da Coruña. Facultade de Informática
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumo]: Actualmente, os ataques de denegación de servizo (DoS) supoñen unha ameaza crecente para a dispoñibilidade de sistemas conectados a internet. Os métodos tradicionais de detección resultan insuficientes ante o volume e complexidade do tráfico de rede.
Este Traballo de Fin de Grao ten como obxectivo deseñar e implementar un sistema de detección de ataques DoS mediante algoritmos de aprendizaxe automática. Para iso, xerouse un conxunto de datos a partir dunha infraestrutura virtual con tráfico lexítimo e malicioso, procesado e empregado para adestrar modelos que permiten identificar comportamentos anómalos de maneira automática. A metodoloxía seguida baséase no proceso CRISP-DM, que inclúe a comprensión do problema, a preparación e limpeza dos datos, a selección e modelado dos algoritmos, e a avaliación do rendemento dos modelos xerados.
[Abstract]: Currently, denial-of-service (DoS) attacks represent a growing threat to the availability of internet-connected systems. Traditional detection methods are increasingly insufficient due to the volume and complexity of network traffic. This Bachelor’s Thesis aims to design and implement a DoS attack detection system using machine learning algorithms. To achieve this, a dataset was generated from a virtual infrastructure with both legitimate and malicious traffic, which was then processed and used to train models capable of automatically identifying anomalous behavior. The methodology followed is based on the CRISP-DM process, which includes understanding the problem, data preparation and cleaning, algorithm selection and modeling, and evaluating the performance of the generated models.
[Abstract]: Currently, denial-of-service (DoS) attacks represent a growing threat to the availability of internet-connected systems. Traditional detection methods are increasingly insufficient due to the volume and complexity of network traffic. This Bachelor’s Thesis aims to design and implement a DoS attack detection system using machine learning algorithms. To achieve this, a dataset was generated from a virtual infrastructure with both legitimate and malicious traffic, which was then processed and used to train models capable of automatically identifying anomalous behavior. The methodology followed is based on the CRISP-DM process, which includes understanding the problem, data preparation and cleaning, algorithm selection and modeling, and evaluating the performance of the generated models.
Description
Editor version
Rights
Attribution 4.0 International







