Brani Vidakovic

Institute of Statistics and

Decision Sciences [ISDS], Duke University


Wavelet-based Approach to Functional Data Analysis

Recent applications multiresolution methods on wavelet-non-standard problems include functional data analysis (FDA). Ongoing research with Gary Rosner from Duke Medical Center is concerned by separating treatment effects in longitudinal observations in a functional ANOVA setup. We present a Bayesian solution of the problem and indicate some other approaches. An example of dimension reduction by wavelets with applications in spectral analysis is also presented.


Back to Colloquium Series