Research

The “Computational Intelligence & Machine Learning” group (CIML) develops basic and applied research on Machine Learning.

The research of the group addresses the development of new Machine Learning methodologies, including  the application to a wide range of innovative Artificial Intelligence (AI) and interdisciplinary fields. In particular,  since the ’90s,  the CIML group has pioneered the development of methodologies for Learning in Structured Domains (sequences, trees, graphs and networks),  the analysis of their computational properties and the  related  applications.

The research is aimed at defining frameworks for the development of intelligent systems and intelligent data analysis tools for learning in complex and structured domains.

The design of new learning methods includes Neural Networks, Probabilistic models, Reservoir Computing, Deep Learning, and Kernel-based approaches, and other Pattern Recognition techniques, with an international leadership for the introduction of efficient Neural Networks and Deep Learning approaches for adaptive processing of structured data.

The application fields include Medicine/Health care, BioInformatics, ChemInformatics, Robotics, Intelligent Wireless Sensor Networks, and Signal/Image Processing. This is particularly important because of the fast growth of information sources due to the explosion in the use of Internet and Sensor Networks and of the new IT capabilities that results from the exploitation and integration of different intelligent approaches.