Section: Partnerships and Cooperations

International Initiatives

Participation In other International Programs

F. Cazals (Inria ABS), I. Emiris (Prof., Univ. of Athens) and S. Theodoridis (Prof., Univ. of Athens) collaborate in the scope of an Inria COLOR entitled Discriminating and classifying in high-dimensional spaces.

The scientific goal was to study methods and algorithms in high dimensional spaces, revolving around three problems: approximate nearest neighbors, polytope volume approximations, and classification - discrimination in high high-dimensional Spaces.

The long-term plan is to examine whether the work done so far can be combined with work by other European teams targeting a European research proposal. F. Cazals and I. Emiris participate in a FET-Open STREP proposal, entitled Exploring the Geometry of Data, including high-dimensional geometry, machine learning, and statisctical methods. More precisely, the collaborations proposed between the two groups bootstraps on the achievements of the COLOR, as they aim at exploring (i) incremental nearest neighbor methods in metric spaces, (ii) sampling methods for polytope volume approximation and high-dimensional space exploration, and (iii) applications in biophysics (protein docking and energy landscape exploration).