Please use this identifier to cite or link to this item:
Title: Aplicación de técnicas de machine learning a la detección de ataques
Author: Rodríguez Rama, José Manuel
Tutor: Hernández Jiménez, Enric
Others: Universitat Oberta de Catalunya
Keywords: computer security
machine learning
Issue Date: 4-Jun-2018
Publisher: Universitat Oberta de Catalunya
Abstract: This project is part of the Final Master Project, specifically the area of "Application of Machine Learning techniques to Security", which I selected as the first option among different possibilities, due to my interest in learning about this subject, which is widely used nowadays. The project is developed in different stages, first using the Weka platform (environment for knowledge analysis of the University of Waikato) and then developing a script written in Python that will do the pre-processing of the dataset, and the subsequent use of a predictive model for detecting malicious connections, specifically using the Scikit-Learn software library. The dataset used is "KDD Cup 1999" which includes a wide variety of simulated network intrusions in a military network environment. This dataset will be analyzed in the present project and will be used to train, test and adjust the selected model. A comparison of various algorithms will be made and applied to the dataset, and the algorithm with best results predicting attacks will be selected, and then tuned to try to find the best configuration using different techniques of Machine Learning and Data Engineering.
Language: Spanish
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
Datasets10porciento.zip4.18 MBUnknownView/Open
TFM.ipynb128.52 kBUnknownView/Open
dec_tree_01.png3.19 MBimage/pngView/Open
jmrodriguez85TFM0618memoria.pdfMemoria del TFM2.71 MBAdobe PDFView/Open
jmrodriguez85TFM0618anexo2.pdfAnexo 2 del TFM133.82 kBAdobe PDFView/Open
jmrodriguez85TFM0618anexo1.pdfAnexo 1 del TFM832.62 kBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons