Brute Force Detection System Based on Machine Learning Classifier Algorithm in Cloud-Based Infrastructure

The increasing adoption of cloud computing across various sectors has led to increased utilization of resources, such as server instances, databases, and microservices. This expansion generates a wide array of log files. The substantial challenge posed by the sheer volume and variety of log files lies in the increasing difficulty of efficiently processing and analyzing them without effective classification. This research focuses on distinguishing brute force attacks from other events in access logs. To achieve this goal, we employ One Hot Encoding for feature extraction and apply machine learning algorithms like Naive Bayes, Decision Tree, Random Forest, and Support Vector Machine. Our findings indicate that Decision Trees and Random Forests are particularly effective, with 87 % accuracy in detecting malicious traffic within log datasets. These results enhance security measures in cloud computing environments and aid in developing more robust and efficient anomaly detection systems.