Dominik Fay

Dominik Fay is a Researcher at the Division of Decision and Control Systems within Kungliga Tekniska Högskolan (KTH). His work focuses on federated machine learning, data privacy, and their applications in healthcare. He is supported by the Wallenberg AI, Autonomous Systems and Software Program (WASP) through their industrial PhD program in collaboration with Elekta, a healthcare technology company.

Education:

• MSc in Computer Science from KTH (2019)
• BSc in Applied Computer Science from Heidelberg University (2017)

His research primarily addresses privacy challenges in distributed machine learning environments. Key contributions include methods for locally differentially private federated learning, dynamic privacy allocation, and privacy amplification techniques tailored for healthcare applications. His work also explores the intersection of machine learning and medical imaging, particularly in segmentation tasks requiring stringent privacy guarantees.

Dominik's publications reflect a strong emphasis on data privacy, federated learning, and healthcare applications. Recent articles focus on correlated noise in federated learning, privacy allocation for composite objectives, and privacy-preserving medical image segmentation. Earlier works extend into smart grid privacy, metabolomics data analysis, and scalable privacy-preserving algorithms.

Grants and Collaborations:

• Supported by WASP Industrial PhD Program
• Collaboration with Elekta, a leader in healthcare technology

He is part of Mikael Johansson's research group, which specializes in decision and control systems, and his work aligns with broader efforts in privacy-preserving AI and machine learning for sensitive healthcare data.