Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity
The Intrusion Detection System (IDS) is an effective tool utilized in cybersecurity systems to detect and identify intrusion attacks. With the increasing volume of data generation, the possibility of various forms of intrusion attacks also increases. Feature selection is crucial and often necessary...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/13/7507 |
| _version_ | 1797592143977512960 |
|---|---|
| author | Pierpaolo Dini Abdussalam Elhanashi Andrea Begni Sergio Saponara Qinghe Zheng Kaouther Gasmi |
| author_facet | Pierpaolo Dini Abdussalam Elhanashi Andrea Begni Sergio Saponara Qinghe Zheng Kaouther Gasmi |
| author_sort | Pierpaolo Dini |
| collection | DOAJ |
| description | The Intrusion Detection System (IDS) is an effective tool utilized in cybersecurity systems to detect and identify intrusion attacks. With the increasing volume of data generation, the possibility of various forms of intrusion attacks also increases. Feature selection is crucial and often necessary to enhance performance. The structure of the dataset can impact the efficiency of the machine learning model. Furthermore, data imbalance can pose a problem, but sampling approaches can help mitigate it. This research aims to explore machine learning (ML) approaches for IDS, specifically focusing on datasets, machine algorithms, and metrics. Three datasets were utilized in this study: KDD 99, UNSW-NB15, and CSE-CIC-IDS 2018. Various machine learning algorithms were chosen and examined to assess IDS performance. The primary objective was to provide a taxonomy for interconnected intrusion detection systems and supervised machine learning algorithms. The selection of datasets is crucial to ensure the suitability of the model construction for IDS usage. The evaluation was conducted for both binary and multi-class classification to ensure the consistency of the selected ML algorithms for the given dataset. The experimental results demonstrated accuracy rates of 100% for binary classification and 99.4In conclusion, it can be stated that supervised machine learning algorithms exhibit high and promising classification performance based on the study of three popular datasets. |
| first_indexed | 2024-03-11T01:48:19Z |
| format | Article |
| id | doaj.art-a0dfcab553bd4a3b93716bdc08f347db |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-11T01:48:19Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-a0dfcab553bd4a3b93716bdc08f347db2023-11-18T16:07:23ZengMDPI AGApplied Sciences2076-34172023-06-011313750710.3390/app13137507Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking CybersecurityPierpaolo Dini0Abdussalam Elhanashi1Andrea Begni2Sergio Saponara3Qinghe Zheng4Kaouther Gasmi5Department of Information Engineering, University of Pisa, 56126 Pisa, ItalyDepartment of Information Engineering, University of Pisa, 56126 Pisa, ItalyDepartment of Information Engineering, University of Pisa, 56126 Pisa, ItalyDepartment of Information Engineering, University of Pisa, 56126 Pisa, ItalySchool of Intelligence Engineering, Shandong Management University, Jinan 250100, ChinaDepartment of the Computer Science, University of Tunis, Tunis 1007, TunisiaThe Intrusion Detection System (IDS) is an effective tool utilized in cybersecurity systems to detect and identify intrusion attacks. With the increasing volume of data generation, the possibility of various forms of intrusion attacks also increases. Feature selection is crucial and often necessary to enhance performance. The structure of the dataset can impact the efficiency of the machine learning model. Furthermore, data imbalance can pose a problem, but sampling approaches can help mitigate it. This research aims to explore machine learning (ML) approaches for IDS, specifically focusing on datasets, machine algorithms, and metrics. Three datasets were utilized in this study: KDD 99, UNSW-NB15, and CSE-CIC-IDS 2018. Various machine learning algorithms were chosen and examined to assess IDS performance. The primary objective was to provide a taxonomy for interconnected intrusion detection systems and supervised machine learning algorithms. The selection of datasets is crucial to ensure the suitability of the model construction for IDS usage. The evaluation was conducted for both binary and multi-class classification to ensure the consistency of the selected ML algorithms for the given dataset. The experimental results demonstrated accuracy rates of 100% for binary classification and 99.4In conclusion, it can be stated that supervised machine learning algorithms exhibit high and promising classification performance based on the study of three popular datasets.https://www.mdpi.com/2076-3417/13/13/7507intrusion detection systemsmachine learningfeature selectiondata managementKDD 99UNSW-NB15 |
| spellingShingle | Pierpaolo Dini Abdussalam Elhanashi Andrea Begni Sergio Saponara Qinghe Zheng Kaouther Gasmi Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity Applied Sciences intrusion detection systems machine learning feature selection data management KDD 99 UNSW-NB15 |
| title | Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity |
| title_full | Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity |
| title_fullStr | Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity |
| title_full_unstemmed | Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity |
| title_short | Overview on Intrusion Detection Systems Design Exploiting Machine Learning for Networking Cybersecurity |
| title_sort | overview on intrusion detection systems design exploiting machine learning for networking cybersecurity |
| topic | intrusion detection systems machine learning feature selection data management KDD 99 UNSW-NB15 |
| url | https://www.mdpi.com/2076-3417/13/13/7507 |
| work_keys_str_mv | AT pierpaolodini overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity AT abdussalamelhanashi overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity AT andreabegni overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity AT sergiosaponara overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity AT qinghezheng overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity AT kaouthergasmi overviewonintrusiondetectionsystemsdesignexploitingmachinelearningfornetworkingcybersecurity |