Contacts

A Physicist in Bocconi

, by Claudio Todesco
Carlo Baldassi is the new assistant professor at the Department of Decision Sciences. His research takes on optimization problems and complex systems, in particular in the field of computational neuroscience, and machine learning

A theoretical physicist doing research on machine learning, Carlo Baldassi could be the personification of the interdisciplinary approach of the Bocconi Institute for Data Science and Analytics (BIDSA). Recently welcomed into the Department of Decision Sciences at Bocconi, coming from the Polytechnic University of Turin, Baldassi firmly believes that big data modeling and management play an important role in a research university. "The amount of data available is enormous and constantly growing. The problem is how to process and use this information. Any institution that starts on this research path holds a huge advantage over its competitors".

Among other things, Baldassi will teach fundamentals of computer science to Bocconi students, before introducing them to more advanced techniques.

Having the strong belief that "theoretical physics is the best way to explore the fundamental laws of nature", Baldassi graduated in 2004 in Trieste. At the International Centre for Theoretical Physics of the city he ran into Riccardo Zecchina, another new recruit of the Bocconi Department of Decision Sciences. He was dealing with abstract optimization problems using the tools of statistical physics. "I was immediately enthusiastic about this amazing approach. It provides powerful research tools to deepen our undressing of seemingly intractable phenomena".

Today, Baldassi's main research areas are optimization problems and complex systems especially in the fields of computational neuroscience and, more recently, machine learning. "The latter is an area where we have made good progress thanks to the use of artificial neural networks inspired by human neural systems. These techniques have an urgent need for in-depth theoretical analysis. It is exciting to understand why they work and how to improve them".

Now the challenge is to keep up with a fast evolving field of research. "The competition is fierce. New studies and outstanding results are published all the time, yet there are still vast, unexplored territories".