Intern
    Data Science Chair

    Recommendation and Security

    In the field of cybersecurity and fraud prevention, our research at the Chair focuses on developing machine learning-based approaches for anomaly detection. We specialize in leveraging advanced techniques to enhance security measures and mitigate risks in various domains. Our primary focus lies in developing and utilizing machine learning algorithms to detect anomalies and suspicious activities in network security, host-based security and fraud detection in enterprise resource planning (ERP) systems.

    We aim to identify potential vulnerabilities and threats within these systems, enabling proactive measures to ensure data and system integrity. Additionally, we delve into the realm of data synthesis, developing methods to generate synthetic data that closely resembles real-world scenarios. To provide transparency and interpretability to our anomaly detection methods, we explore Explainable Artificial Intelligence (XAI) techniques. These methods enable us to offer insights into the decision-making process of our machine learning models, aiding users in understanding the rationale behind anomaly detections and gaining insights into potential security threats.

     

    Core Research Topics include:

    • Explainable Artificial Intelligence
    • Anomaly detection
    • network flow data generation

    Projects

    • Occupational Fraud Detection - Detecting fraud within the large amount of data tracked by companies through Enterprise Resource Planning (ERP) systems.

    • DZ.PTM - Recommendations for clinical diagnostics in cooperation with the University Clinic Würzburg.

    • HydrAs - Developing methods for Hypothesis-driven Analysis of Sequential Data. 

    Concluded Projects

    • DeepScan  - Machine Learning for fraud detection in enterprise resource planning software systems.
    • adidas - Modeling and Personalization for Customer Engagement

    Publications

    Here is a list of selected publications.