SpreadSentinel introduces an incremental framework for early DDoS detection in dynamic, imbalanced network environments, combining forward-chaining techniques, mini-batch learning, and SMOTE-based class rebalancing. Evaluated on the CICDDoS2019-Friday dataset, the approach surpasses 99% detection accuracy and is adaptable to SDN and IoT infrastructures. Best Paper Award, PA-AISS Conference 2025.

A behavior-based approach for detecting obfuscated Android malware, combining model-based and AI-based techniques to observe malware behavior indirectly through the runtime operations of Android services rather than attempting deobfuscation.

A middleware architecture for Android malware detection using machine learning and cybersecurity-oriented mobile analysis.

An investigation toward improving DeepMalOb through formal security risk analysis methods for Android malware detection.

A further investigation of UFILA for Android malware detection through feature updates using incremental learning.

An empirical study of how imbalanced Android malware datasets affect the efficiency of machine learning models.

A zero-day Android malware detection approach using deep learning and zero-shot learning techniques.

A deep learning approach for detecting obfuscated Android malware using memory dump and malware analysis features.

A best-paper awarded contribution on updating Android malware classification features through an incremental learning approach.

A study of the impact of imbalanced datasets on machine learning performance for Android malware detection.

A preliminary investigation of mobile banking attacks in West Africa based on feedback from Orange Money customers in Burkina Faso.

A blockchain-based approach for secure, transparent, and uniform mobile money in Internet-underserved areas.

A return on experience on bootstrapping software engineering training in developing countries at the University of Ouagadougou.