Home People Picci Biography

Biography

Giorgio Picci received the Dr. Engineerig degree (cum laude) in Electronic Engineering,
   from the  University of Padova, Italy, in July 1967.       Since 1980 he is Professor with the      Department of Information Engineering         University of Padova, Padova, Italy.
 Chair: Model Identification.
Previous Positions (in Italy):
      November 1970--October 1975: Lecturer with the
Department of Statistics, University of Padova.
August 1973--October 1980: Researcher with 
LADSEB-CNR,  Laboratory of the National Research
Council, Padova, Italy.
 November 1975--October 1980: Associate Professor,
Institute of Electrical and Electronic Engineering,
University of Padova .

Extramural Appointments:

February--October 1973: Visiting Assistant Professor, Division of Applied Math., Brown University, Providence, R.I. USA.

Spring Semester 1976: Visiting Scientist, Electronic Systems Laboratory, M.I.T.,  Cambridge, Massachussetts. USA.

Spring Semester 1980 and Fall Semester 1982:  Visiting Associate Professor, Department of Mathematics, University of Kentucky, Lexington, Kentucky. USA

Academic Year 1986/87: Visiting Professor, Department of Electrical and Computer Engineering, Arizona State  University, Tempe, Arizona, USA.

Fall Semester 1988: Distinguished Visiting Professor, Department of Electrical and Computer Engineering, Arizona State University, Tempe, Arizona, USA.

September-October 1994: Visiting Scholar, Dept of Applied Mathematics and Physics, Kyoto University, Kyoto, Japan. Academic year 1995/96: Visiting Professor, Department of System Science and Mathematics, Washington University, St. Louis, Missouri.

Many one-month appointments and shorter visits  with the  Division of Optimization and  System Theory, Department of Mathematics, Royal  Institute of Technology, Stockholm, Sweden  1985--present.
February 2003: Invited visiting professor at the Mittag-Leffler Institute, Stockholm, Sweden.

Professional Interests:

Professor Picci's most outstanding contribution has been in the area of  modelling, realization and identification of Stochastic Systems.  This research  has made it possible to build a  geometric framework for the study of Stochastic System centered around the idea of  Markovian Splitting Subspace which is a natural probabilistic analog of the  notion of   state   and state--space model in deterministic system theory. The geometric approach has led for the first time to the  understanding of structural

properties of stochastic models like for example  minimality  which are important in applications, e.g. for characterizing the minimal dimension  of estimation algorithms and solvability  of various linear estimation problems.

The new geometric framework has led for example to an elegant new  solution of a class of noncausal estimation problems and to elucidate the  role of the family of   all solutions of the Algebraic Riccati Equation in computation of the optimal noncausal estimator. It has also provided a unified  theoretical background for the family of identification methods now  known as subspace methods. Within  this framework a unified analysis of these methods has been carried on leading for the first time to understand the  common features of many seemingly unrelated algorithms proposed in the literature  and to a transparent assessment of the statistical properties of these estimators.