Title of the Talk: 

Designing Federated Learning Frameworks for Secure Pharmacogenomic Databases

Anstract:

Pharmacogenomic data plays a vital role in personalized medicine, but its sensitive nature raises significant privacy and security concerns. This study proposes a federated learning framework that enables multiple healthcare institutions to collaboratively train models without sharing raw patient data, reducing data exposure risks by over 70%. The architecture is designed to handle large-scale genomic datasets (exceeding terabytes) while maintaining high model accuracy (above 90%) and compliance with data protection standards. Experimental evaluations demonstrate improved data security and a 30–40% reduction in data transfer requirements compared to traditional centralized approaches. The framework offers a scalable and secure solution for advancing precision medicine.

Profile:

Database Administrator based in Linden, New Jersey, with strong expertise in SQL, big data technologies, and statistical data analysis. Currently pursuing a PhD in Information Technology (expected 2027), with a focus on healthcare capstone research. Skilled in database management, technical support, and data-driven problem solving, with a commitment to delivering efficient and reliable data solutions.