Yuhong Yang thinks his NSF CAREER Award will aid numerous other
academic disciplines.
When it comes to statistical data, Yuhong Yang, assistant professor of
statistics, says random errors do occur.
He'll now have the opportunity to research those errors and find solutions
that will be applicable to numerous other academic disciplines.
"In applications, a difficulty a user often faces is the choice of
the best method to be used for the data at hand," Yang said.
This is especially the case for high-dimensional function estimation, where
to overcome what Yang refers as "the curse of dimensionality,"
various methods have been and/or will be proposed according to different
characterizations of the target function.
Yang is striving to construct adaptive estimators by combining a collection
of candidate procedures.
"The goal for the combined procedure is to perform automatically as
well as, or nearly as well as, the best original procedure without knowing
which one it is," he said.
What makes Yang's research more interesting and challenging is that he will
deal with generally dependent random errors instead of independent or short-range
dependent errors as is usually assumed. Long-range dependence refers to
the case when the correlation between two errors remains strong even when
two observations are far part either in space or in time.
Long-range dependencies have been observed in many applied scientific disciplines
including astronomy, physics, geoscience, hydrology and signal processing.
Statistical issues become much more complicated when long-range dependence
happens.
Yang's research in this area will now be funded by a National Science Foundation
(NSF) CAREER Award from the NSF's Division of Mathematical Sciences. The
five-year $250,000 grant is entitled "Adaptive Regression for Dependent
Data by Combining Different Procedures."
The statistics professor has been researching this phenomenon since he came
to Iowa State.
"This grant was a natural extension for me to continue work in this
area," he said. "My strength has been on the theoretical side
of statistics.
"Some researchers in the field are aware of this problem," Yang
continued. "Many are not however. We want to work with other disciplines
and discuss the issues that arise from this and provide appropriate statistical
solutions."
Yang plans on collaborating with Iowa State professors and research groups
in some academic disciplines including atmospheric science, electrical engineering,
and agronomy among other areas.
"I hope to work with several other groups on campus," he said.
"We are already working with a group in electrical engineering. The
more we work with these groups the more questions we will be asked and will
be asking. By answering those questions, we'll be able to provide these
groups with methods that will make them more efficient from a statistical
point of view.
"There's a lot of opportunity in this area in the future," he
continued. "But I know that by the end of the grant's funding, many
of these problems will be solved."
