KAUST Research Workshop on Optimization and Big Data
University of Oxford
Raphael Hauser is a faculty member in the Numerical Analysis Group at the Mathematical Institute at the University of Oxford in the United Kingdom. His research interests focus on convex optimization and convex analysis. On the theory side he is interested in complexity theory and the probabilistic analysis of algorithms, as well as in condition numbers and the design of algorithms, particularly distributed algorithms for high dimensional problems. On the applications side, Prof. Hauser is interested in the design of optimization models in mathematical finance, in medical imaging and dimensionality reduction for data science.
A new generation of 3D tomography systems is based on multiple emitters and sensors that partially convolve measurements. A successful approach to deconvolve the measurements is to use nonlinear compressed sensing models. We will discuss two different nonlinear compressed sensing models and algorithms to deconvolve the measurements in a high dimensional setting, resulting in 3D image reconstruction.