ETD

• anomaly detection
• cybersecurity
• deep learning
• explainable ai
• machine learning
• time series

This thesis project was developed during a curricular internship, addressing challenges in modern cybersecurity through an innovative approach to anomaly detection in time series analysis, with a significant implementation of explainable AI methodologies. The work represents a unique intersection between industry-oriented research and academic investigation, bridging practical cybersecurity applications with theoretical advancements in machine/deep learning interpretability. The research contributes to the field by proposing novel techniques for detecting anomalous patterns in temporal data while maintaining transparency and interpretability in the decision-making process, which is crucial for cybersecurity professionals who need to understand and validate automated threat detection systems in real-world enterprise environments.