Interview with Alumni Award Winner Dr. Piercesare Secchi
Distinguished Alumni Award winner Piercesare Secchi (PhD, ‘95) looks back on his graduate school experience at the U with great fondness, and remains close friends with fellow graduate students and professors from that time. Nowadays, he serves as a Professor of Statistics at the Department of Mathematics at Politecnico di Milano, which he led from 2009-2016. While challenging, Secchi describes the demands of his work there as gratifying. He is quick to acknowledge how others have helped him succeed in his endeavors and share his insight on the future of the field.
What brought you to the U of M School of Statistics for your degree? How did you end up studying statistics?
In February 1989 I was a doctoral student at the University of Trento, in Italy, pursuing a degree in Statistics, after a Laurea in Mathematics from the University of Milan. At that time, I attended a conference on subjective probability at Erice, in Sicily, and there I met professors Bill Sudderth and David Lane, both of them at the U of M School of Statistics. I knew them already as authors held in high esteem by my Italian advisor, professor Eugenio Regazzini. It was then natural to approach them during a conference dinner. This was a turning point in my life as a statistician, since at the end of the dinner David Lane proposed that I visit the School to pursue my research on subjective probability and Bayesian statistics. In August 1989, with my wife Elena Freyrie - married a month before! - I reached Minneapolis and I enrolled in the PhD program of the School. Eventually I ended up with two doctoral degrees in statistics, one from the University of Trento and the other from the U of M!
What is one of the defining experiences from during your time as a statistics student that you still think about?
It’s quite difficult to single out a single moment of an extraordinary experience that lasted for a few years – too few! – and which is now narrated, by my wife and me, as the creation myth of our own family! My memories of those wonderful years are populated by a lot of friends, most of them still very close. Some were fellow graduate students, some were my professors at that time. And then of course the U of M, Vincent Hall – where the stat department was at that time – and then Minneapolis, with its lively cultural and social life. A city that we loved - although it’s very different from Milano!
What do you do now? Where do you currently work?
I am now Professor of Statistics at the Department of Mathematics of Politecnico di Milano. Although I began my career as a researcher working on stochastic games with Bill Sudderth and nonparametric Bayesian statistics with Pietro Muliere, both of them among my best friends, in the last 20 years my interests moved more and more toward real world applied statistics problems, stimulated by interactions and projects developed with colleagues in engineering and computer science at Polimi, which many regard as the top technical university in Italy. My current research is focused on methods and algorithms for functional data analysis and object oriented spatial statistics.
In addition to completing significant research in many areas of probability, game theory, and statistics, you teach and have served as chair of one of the top math departments in Italy. How do you do it all?
Well, my eight years serving as Head of the Department of Mathematics of Politecnico di Milano were a real challenge! But also, a very gratifying experience. The department is big - counting a faculty of more than 100 people, plus doctoral students, teaching assistants, administrative staff - and offers all courses in mathematics, statistics, and mathematical finance taught in a technical university with more than 40,000 students, in engineering, architecture and design. Honestly, pursuing scientific research while coping with academic politics was a demanding job, but I was helped all along by an exceptional group of junior colleagues and doctoral students working with me in the Stat Group at MOX, the modeling and scientific computing laboratory of the Polimi Department of Mathematics.
How has statistics changed since you started your work in it? What should current students understand about your profession if they are considering it for themselves?
The current hype on big data, machine learning and data science is both a blessing and a curse for any statistician of my generation. The passionate interest in data science of so many and heterogeneous industries is attracting more bright young students to academic curricula where statistics plays a major role. Research, both methodological and applied, on statistical learning receives renewed attention and financial support from institutional research agencies and from tech companies. These are opportunities to advance our discipline that, as academics and professional statisticians, we should not miss. There is however a price to be paid; we need to be open to revise our approach to data analysis along the lines already indicated by Leo Breiman in his 2001 seminal paper on the two cultures in statistical modeling. We should not be prejudiced about the machine learning and data science achievements fostered by our colleagues in computer science; on the contrary, they are encouragements to revise our established paradigms, they must drive our energetic effort to explore their connections with our “received” statistical thinking while founding them on novel mathematical arguments.
What do you like most about being a statistician?
On the door of my office, I posted the centerfold of an old AMSTATNews issue with John Tuckey’s words: “the best thing about being a statistician is that you get to play in everyone's backyard.” Working in a technical university gave me a lot of opportunities to cultivate my curiosity for science driven by data – a curiosity which I believe is shared by most statisticians – and many backyards where to play with statistics!