Efficient Hybrid Model for Botnet Detection Using Machine Learning

In recent years, with the advancements in technology, threats to internet security have also been increased and botnets are among those threats. Botnet is a very critical and important area of research in the field of cybersecurity. Botnets are basically a network of computers or other internet-connected devices that are infected by malware and controlled by a cybercriminal. They are used for various malicious purposes like spamming, phishing, DDoS (Distributed Denial of Service) attacks, and so on. One of the most dangerous uses of botnets involves perpetrating fraud which can even lead to financial losses for individuals as well as organizations. Detecting botnet fraud is a challenging task because of its complexity and sophistication. Therefore, it requires a high level of research for its detection. In the current research paper, ML (Machine learning) algorithms have been used for botnet fraud detection. A hybrid model (KNN, Random Forest, and Logistic regression) is built on the dataset which has yielded 86.97 percent accuracy, 81.76 percent precision, 98.25 percent recall, and 89.25 percent F1-score in detecting botnet attacks. We have also discussed the challenges associated with botnet detection and have identified potential future directions for research in this area.