Functional Data Analysis
- I work in the area of functional data analysis (fda)
- It is the analysis of a data set that contains functions (curves)
- Many studies contain data that is measured repeatedly over time (or space) for each unit. This type of data can be viewed as a function of time (or space)
- Anytime data is collected over time/space fda comes handy
- Examples: Finance, health, climate studies, genetic studies, imaging studies (fmri etc.)
some Recent work
Association studies
- As the name suggests association studies determine whether wo variables are associated with each other
- My interest is to develop association detection methods when a functional variable is involved
- Link: Kendall’s Tau for functional data analysis
Clustering
- How do we cluster a bunch of curves?
- Link: Pan-Disease Clustering Analysis of the Trend of Period Prevalence
Measurement Error
- How do we get “good” estimates in a regression model involving functional data, if the functional data is contaminated with erros?
- Link: Functional measurement error in functional regression
curve registration
- How do we get “good” estimates in a regression model involving functional data, if the functional data is contaminated with erros?
- Link: Functional measurement error in functional regression
deep learning and functional data analysis
- using neural networks for prediction in functional regression model
- using neural networks for outlier detection in functional data
- using ideas from functional data analysis to add flexibility to sequential neural networks