RNN-based Malicious DNS Sequence Detector (SPL Project) — In Progress
Developed an RNN-based malicious DNS sequence detector as an ML project to support cybersecurity use cases. In this work, model development required preparing and using DNS traffic sequences for supervised learning and classification of malicious vs benign patterns. The project focuses on training on DNS traffic datasets to learn anomaly characteristics from sequence data. • Implemented core neural network logic in C++ for efficient sequence classification and anomaly detection. • Used DNS traffic sequence datasets to capture malicious query patterns. • Trained and evaluated an RNN model for sequence-level anomaly detection. • Targeted real-world cybersecurity applications involving DNS query behavior modeling.