Research
In general, people at BESTA are interested in analyzing high-frequency, massive, bio related data and applications of wavelet and other dimension-reducing linear and non-linear transformations to the measurements. The current research projects include:
- Scaling, Fractal and Multifractal Measures in 1-, 2-, and 3-D data. Corresponding spectral tools and summaries. Use of spectral descriptors in inference.
- Bayesian Statistics: Bioengineering Applications in Meta Analysis and Adaptive Designs. Prior Elicitation. Bayesian Nets.
- Multiresolution enhancement of signals and images.
- Accounting for Bias in Health-related reporting.
- Statistical designs when the observations are functions and images. Functional Data Analysis.
- Wavelets on spheres and topologically equivalent manifolds. Regularization techniques.
- Semi-supervised learning. Incorporation of information carried by ``unlabeled'' items to learning process supervised by the ``labeled'' items. Closeness measures on manifolds and Laplacians.
For more details see BESTA's poster presented at 10th aniversary of BME, March 2008: [ POSTER: An Overview of Current Projects ]
