Parallel Scientific Computing Laboratory Group (PSCL)

The PSCL research fields mainly focus on the development of parallel algorithms for scientific problems with a current emphasis on the use of machine learning techniques.

Our aim is the study of problems of the real world through simulation. So we develop parallel algorithms and software for the prediction of complex physical phenomena (e.g weather, ocean, air pollution, fire trajectory). In combination with the quest of parallel software development, domain decomposition methods and load balancing algorithms are also under study. Another field of interest is the study of problems in the area of combinatorial scientific computing. Especially, we study the numerical solution of Laplacian systems which arise from graph problems (e.g Page Rank). The PSCL was founded in 1995 by Professor Nikolaos Missirlis.

Scientific computations for forecasts graphic
The usefulness of scientific calculations in solving problems related to forecasting of:  the weather  gas pollution  the ocean (tsunami behavior)  cyclone orbits  earthquakes  stock market behavior, etc. Researchers: Prof. N. Misyrlis, Assistant Professor F. Tzaferis, Dr Maria Louka (…