Signals and Learning Group

We work on the intersection of machine learning, signal processing and system theory: advancing fundamental theory and translating it into new applications, with particular focus in medicine. List of current members:

• Postdoc: Elis Stefansson. Complexity and robustness in learning and decision-making.
• Postdoc: Petrus Abreu. Automatic ECG analysis and telehealth. * (main advisor: Antonio Luiz P. Ribeiro)
• PhD: Han Wu. Large-Scale Robust Machine Learning via Minimax Formulations.
• PhD: Jiawei Li. Foundation models for electrocardiography.
• PhD: David Väviggreen. Robust methods for unsupervised machine learning.
• PhD: Bror Hultberg. Anytime inference with application to precision medicine. * (main advisor: Dave Zachariah)
• PhD: Karthik Danda. Theory and methods for distributed continuous learning. * (main advisor: Ayca Özcelikkale)
• PhD: Sofia Falco. Agentic AI in telehealth. * (main advisor: Frederico Guimaraes Gadelha)
• MSc: André Ramos Ekengren. Methods for Combining Patient Characteristics and Signal in AI-ECG.
• Visiting Postdoc: Pâmella Miranda. Oportunistic screening for neglected tropical diseases.. (UFMG, Brazil).

Alumni:

• PhD
  • Daniel Gedon. On Deep Learning for Low Dimensional Representations (Defence: June 2024).
• M.Sc.
  • Sabereh Hassanyazdi. Machine Learning for Modeling Breathing Signals and Predicting Physiological Markers (Defence: June 2026).
  • Wenhan Zhou. Uncertainty Quantification for Receiver Operating Characteristic Curves (Defence: June 2025).
  • Arvid Eriksson. Adversarial training for the diagnosis of the ECG (Defence: June 2024).
  • Oscar Larsson. Generation and Detection of Adversarial Attacks in the Power Grid (Defence: July 2022).
  • Theogene Habineza. Deep Learning-Based Risk Prediction of Atrial Fibrillation Using the 12-lead ECG (Defence: June 2022).
  • Samiha Akhter. Reinforcement Learning in Large Language Models (Defence: June 2026).
  • Aditya Khadkikar. Federated Machine Learning Techniques for Automated ECG Classification (Defence: June 2026).
  • Olle Wernersson. Interpretable deep-learning models for automatic ECG classification (Defence: June 2026).
  • Johan Millberg. Evaluating clusters of financial time series (Defence: June 2023).
  • Christie Courtnage. An extension to Semi-Supervised Learning using Shapley Value Data Valuation (Defence: June 2022).
  • Meenal Pathak. Automated Accounting using Machine Learning (Defence: Apr. 2022).
• Visiting students
  • Andrea Basteri. Multiple imputation and Online EM. Until June 2026 (INRIA, France).
  • Victoria Sitati. Interpretability and repeatability in AI-ECG diagnosis (CORE-AI program). Until Nov. 2025 (DeKUT, Kenya).
  • Kellen Sumwiza. AI-ECG age in the prediction of cardiovascular diseases (CORE-AI program). Until Nov. 2024 (University of Rwanda, Rwanda).
  • Lennert Bontick. AI-ECG invariant to patient Characteristics via Gradient-Reversed. Until Sep. 2024 (Ghent University, Belgium).
  • Karel Fonteyn. Orthogonalized deep learning for ECG Risk Prediction. Until Sep. 2024 (Ghent University, Belgium).