A cura di MIUR - Direzione Generale per il coordinamento la promozione e la valorizzazione della ricerca

Semeion, mathematics to understand complexity

Semeion, mathematics to understand complexity

“We all use mathematics every day: to predict the weather, to tell the time, to count the money. We even use mathematics to analyse crimes, understand patterns, and predict behaviours. Using numbers, we can reveal the greatest mysteries of life!" These are the words of Charlie Eppes on Numb3rs, CBS's hit television series in which we can appreciate the endless applications of mathematics in our daily lives.

However, you need not go far to find research centres of excellence. Even here, researchers work with numbers and formulas to understand phenomena of incredible complexity, able to move from the highest level of abstraction to more concrete practice, to give solutions to problems still considered impossible to solve, such as predicting natural disasters, early diagnosis of neurodegenerative diseases, and discovering the home of an "unknown subject" who has just committed a crime.

One of these centres is Semeion – Research Center of Sciences of Communication in Rome. Since 1985 this centre has been conducting basic, experimental and applied research in the field of artificial intelligence. Today the Special Scientific Institute of the Ministry of Education, University and Research (MIUR), Semeion is closely involved with several Italian and international institutions including the National Research Council (CNR), the National Institute of Nuclear Physics (INFN), the Italian Space Agency (ASI), the National Institute of Geophysics and Volcanology (INGV) as well as with the University of Colorado, the United Nations Development Programme (UNDP) and Scotland Yard.

Over the years, at the Semeion laboratories the researchers have designed, engineered and tested new models of "artificial adaptive systems", i.e. neural networks, evolutionary algorithms, data analysis and pre-processing systems as well as computer programmes aimed at testing mathematical models. The research has resulted in over 200 publications of scientific articles on international journals, several monographs and 11 international patents which cover four main application areas: social security and crime; biomedical diagnostic imaging and clinical pharmacology; economic and financial forecasts; land-use planning and management.

buscema_700

"At Semeion we study complex systems, those systems that create their own operating rules: these rules cannot be known at the beginning of the process as they are generated by the interaction among the various parts," explains to ResearchItaly Massimo Buscema (picture), director of the centre and professor of Mathematics and Statistics at the University of Colorado in Denver. "In this case - continues Buscema - you must proceed by taking the minimum pieces of information from the system and let them interact until the intrinsic model of the process emerges from the bottom. It is just like a film: to understand it, it cannot be reduced to a single picture or a single frame."

Everything comes from the world of artificial intelligence that historically develops three complementary approaches: the simulation of the human brain to understand how it works (bioengineering); the emulation of the human brain to build models with similar function (computer science); the physical approach "which is inspired by the functioning of the brain as a material artefact to understand the laws underlying the change of individual behaviour into collective behaviour, such as the physics of gases, aggregations of molecules, the crowd and swarm dynamics." The last one is the preferred approach at Semeion Rome.

folla_700

Source Redazione ResearchItaly
Publication date 04/20/2015