|
" Detection of Intrusion Through Utilizing Machine Learning Algorithms "
Oztemel, Muhammed Esad
Salam, Mohammad A.
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1105094
|
| Doc. No
|
:
|
TLpq2296357282
|
| Main Entry
|
:
|
Oztemel, Muhammed Esad
|
|
|
:
|
Salam, Mohammad A.
|
| Title & Author
|
:
|
Detection of Intrusion Through Utilizing Machine Learning Algorithms\ Oztemel, Muhammed EsadSalam, Mohammad A.
|
| College
|
:
|
Southern University and Agricultural and Mechanical College
|
| Date
|
:
|
2019
|
| student score
|
:
|
2019
|
| Degree
|
:
|
M.S.
|
| Page No
|
:
|
80
|
| Abstract
|
:
|
The importance of the network security has been dramatically increasing since the internet usage plays an important role in our daily life. In this study machine learning based network intrusion detection system (NIDS) have been developed to provide a secure infrastructure for users. For the proposed NIDS Deep, Convolutional Neural Network (DCNN) as well as different machine learning algorithms such as K-Nearest Neighbor (KNN), AdaBoost, Decision Tree, Naive Bayes, Multilayer Perceptron have been explored to classify the network traffic detection. In this study two models have been developed to detect the network intrusion. The first model has been developed to classify the network traffic as attack or benign. On the other hand, the second model classifies algorithm the attack types as well as benign traffic. Publicly available CICIDS2017 dataset has been used in this study for training and testing. In the selected dataset, there are eight different types of networking attack alongside benign traffic. According to the results, all selected algorithms have provided more than 90 % overall accuracy performance for model 1 and model 2. Furthermore, DCNN has provided the best performance and Naive Bayes has provided the lowest performance within the selected algorithms.
|
| Subject
|
:
|
Computer science
|
| |