The research conducted by the staff of the laboratory is related to the following topics.
Machine learning and data mining
- Incremental learning of prediction models from concept drifting evolving data streams
- Improving classification of class imbalanced data
- Integrating symbolic knowledge into neural networks
- Applications of machine learning and data mining methods to biomedical data
- Mainly supporting diagnostic and therapeutic decisions, medical imaging, improving patient adherence to therapy
- Learning from complex data
- Online learning from complex representations
- Handling incomplete and delayed information
- Complex instance representations and structured output
- Text mining and natural language processing
- Learning from and with structured data and knowledge representations
Trustworthy AI and explaining ML models
(Research, design, and develop a unified methodology for interpretable machine learning systems, especially for areas where close interaction between humans and technology are essential).
- New learning architectures
- Evaluation criteria and frameworks
- Preference modeling and learning
- Perception and human in the loop
- Visualization
- New modes of interaction with users/humans
- Responsible AI and fair decision making
Other research areas
- Process mining – analysis of event logs for process modeling
- Algorithms for exact and heuristic optimization
- Spiking neural networks (SNN) and signal analysis
- SNN learning algorithms for edge applications
- Deep learning for natural language processing
- Neurosymbolic learning and reasoning
- Multi-agent systems, evolutionary robotics, biological simulations, artificial life
- Optimizing topologies of neural networks for real-time control
- Muliticriteria decision analysis