Federated Learning for Private, Fair, and EfficientAI for Well-being

Important results from the project

The project had two main objectives: 1) to produce at least two high-quality papers, and 2) to establish a long-term collaboration with the Aiim lab at Queen´s University in Canada. We are currently working on two papers, with plans to submit to ICML and NeurIPS 2025. Publishing high-quality papers in just 4 months is challenging, but we will continue toward this goal. As for the second objective, we initiated discussions on future international projects, to be able to continue our work on Federated Learning using Time Series data, achieving the project´s overall goals.

Expected long term effects

Detta projekt resulterade i ett samarbete med Queen´s University i Kanada, med fokus på att utveckla metoder för Federated Learning med ett primärt fokus på välbefinnande. Vi analyserade tidsseriedata, vilket ledde till två artiklaren om analys och en om metodologisom är tänkta att skickas in till ICML 2025 och NeurIPS 2025. Projektet hade två huvudmål: att producera högkvalitativa artiklar och att etablera ett långsiktigt partnerskap med Aiim-labbet. Vi gör framsteg inom båda, inklusive att initiera diskussioner för framtida internationella projekt.

Approach and implementation

One aspect of this project, despite being a mobility project, was overcoming the challenges related to travel due to visa issues. We successfully initiated our collaboration remotely, holding weekly meetings with the main supervisor from the Aiim lab (Prof. Ali Etemad). Additionally, I became involved in their research group through the Slack application, allowing us to share knowledge and stay connected. We are committed to continuing our work on the papers, not only under Prof. Etemad´s guidance but also in collaboration with his team.