The research has been focused on distributed algorithms for modelling and optimization of adaptive systems.
In particular, neural and neuro-fuzzy paradigms have been refined for fast and accurate complex system identification, classification and pattern recognition. These tasks usually require huge amounts of data and computations. Therefore, innovative granular computing techniques and evolutive algorithms have been developed to support semantic segmentation, clustering, classification of multidimensional data (still images, audio and video sequences).
- Granular computing
- Independent Component Analysis
- Neural networks
- Neuro-fuzzy networks
- Pattern recognition
- Probabilistic optimization