Statistical insight with Functional Data Analysis
Functional Data Analysis is a branch of statistics that emerged out of nonparametric statistics to study data where one or more variables can be viewed as a curve, surface, or other type of function.
Penn State is home to several researchers making pioneering developments in functional data analysis including Drs. Francesca Chiaromonte, Bing Li, and Matthew Reimherr. Their contributions have covered a number of areas in Functional Data including predictive modelling, dimension reduction, data privacy, alignment, clustering, motif discovery, independent component analysis, and statistical graphical modelling. They are also involved in collaborative biomedical applications utilizing functional data analysis techniques.
The impacts of their biomedical collaborations are captured in these recent articles:
Young children’s oral bacteria may predict obesity
Unusual DNA folding increases the rates mutations
Staying home, primary care, and limiting contagion hubs may curb COVID-19 deaths
Using curves to analyze multifaceted data may hold the key to understanding childhood obesity
Dr. Li has authored a book on Sufficient Dimension Reduction, that includes sections on functional data, while Dr. Reimherr co-authored an introductory book on Functional Data. They have received several national research grants to support this effort and their students have gone on to take jobs in a variety of areas in academia and industry.