IMPROVING INTRUSION DETECTION USING TREE ADJOINING GRAMMAR GUIDED GENETIC PROGRAMMING

• Main Authors: Vũ Văn Cảnh, Hoàng Tuấn Hảo, Nguyễn Văn Hoàn
• Format: Article
• Language: English
• Published: Dalat University 2017-09-01
• Series: Tạp chí Khoa học Đại học Đà Lạt
• Subjects: lập trình gen; phát hiện xâm nhập; phân loại tấn công; văn phạm nối cây.
• Online Access: http://tckh.dlu.edu.vn/index.php/tckhdhdl/article/view/339

Nowadays, the problem of network security has become urgent and affect the performance of modern computer networks greatly. Detection and prevention of network attacks have been the main topic of many researchers in the World. One of the safety measures for networks is using the intrusion detection systems. However, these measures are costly, ineffective, unreliable and can-not detect new or unknown attacks. Some studies using machine learning technology have been applied in intrusion detection. In our work, we proposed using Genetic Programming (GP) to improve intrusion detection. In the experiments, we used GP and Tree Adjoining Grammar Guided Genetic Programming (TAG3P) on artifical datasets suggested by Pham, Nguyen, and Nguyen (2014). Compared with previous results, we found that GP and TAG3P are more effective in detecting attacks than previous measures.