INTELLIGENT ALGORITHMS FOR DETECTING NETWORK PORT ATTACKS IN CYBERSECURITY

This study is devoted to the problem of intelligent detection of network port scanning attacks, which is an important aspect of cybersecurity. The paper analyzes the limitations of traditional signature-based systems, particularly their inability to effectively detect low-intensity and stealth attacks. A hybrid model combining statistical methods (Shannon entropy) and machine learning (Random Forest) is proposed. The proposed approach analyzes network traffic in real time, ensuring high efficiency and a low error rate. The practical part of the study is implemented in the Python environment and validated using the CIC-IDS2017 dataset.