The outcome of this project is a web application and a research paper titled "IOTs Traffics Detection and Analysis using Machine Learning for Cybersecurity Application" and published in 2023 IEEE 5th Eurasia Conference on IoT, Communication and Engineering (IEEE ECICE 2023) URL: https://www.ecice.asia/ National Formosa University, Yunlin, Taiwan - on 27-29 October 2023 Organized by:
- College of Engineering, National Formosa University, Yunlin, Taiwan
- Institute of Electrical and Electronics Engineers (IEEE)
- International Institute of Knowledge Innovation and Invention (IIKII)
- Smart Machinery and Intelligent Manufacturing Research Center
In this research, feature extract technique is applied to detect and analyze IoTs Benign and Attack traffics’ features from up-to-date and large-scale dataset called CICIoT2023 from Canadian Institute for Cybersecurity - University of New Brunswick. The dataset contains a diversity of traffic data types generated by IoT lab connected 105 smart devices, also the selected features are then applied to several machine learning algorithms in order to understand the IoTs traffic behaviors for better security applications and analytics. The applied algorithms have achieved satisfactory to high performance. The highest F score results were Decision Tree 0.979, KNN 0.973, Naive Bayes 0.704, Random Forest 0.939, MLP 0.902.
Additionally, I have developed a simple web application using Python to easily upload a dataset, visual data, and train ML models. The used code is shared in other files. To run this application, you need to apply your dataset, use Streamlit Python as well. The application Main page is:
commands needed: streamlit run "codefilename.py"