Work

Effiziente Federated Learning-Systeme in ressourcenbeschränkten Umgebungen unter neuer KI Gesetzgebung

Year: 2025

Type: dissertation

Abstract: This thesis examines the challenges and opportunities of implementing Federated Learning (FL) systems on embedded hardware, with a focus on scalability, efficiency, and regulatory compliance. We first perform hardware-centric studies including tiny models (< 100K model parameters) and LLMs, where we identify opportunities to jointly optimize computation and communication efforts. We complement our work with an in-depth study of the EU AI Act and its effects on FL.

Source: mediaTUM (Technical University of Munich)

Author: Woisetschläger, Herbert

Language: English

Topic: Privacy-Preserving Technologies in Data

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funders: Deutsche Forschungsgemeinschaft, Bayerische Staatsministerium für Wirtschaft, Landesentwicklung und Energie

OpenAlex ID: W7114825201

Cites: 140

Cited by:

Related to: 10