The U.S. Department of Energy has asked Giovanni Petris, an associate professor of mathematical sciences at the University of Arkansas, to collaborate in the
development of effective statistical models to forecast energy prices
for various energy sources in the next decades.
With the recent and current
development of new technologies for the extraction of oil and gas, the
relationships among the prices of different energy sources — as well as
the relationships linking those prices to other fundamental economic
variables — are likely to undergo profound structural changes whose
nature is still highly uncertain.
It is therefore of paramount
importance, in order to obtain useful forecasts, to employ statistical
models that incorporate experts' input but are at the same time
flexible enough to quickly adapt to the evidence provided by new data
as these become available.
Professor Petris is an expert in time series analysis and a renowned leader
in the field of dynamic models, which are particularly well-suited for
the prediction problem the Department of Energy is facing.
In addition to the many
research articles in this area, he has developed a software package
and written a book on the subject. Both the software, written in the
popular statistical programming language R, and the book are used by
researchers, practitioners and students worldwide.