Applying Artificial Intelligence to Predict the Performance of Data-dependent Applications

Computational science (CS) is an emerging discipline that unites science and mathematics with disciplinary experience in biology, chemistry, physics, and other applied scientific fields. Within the scientific method, it is often referred to as the third science paradigm, complementing both theoretical and laboratory science (Miller & Boxer, 2005). In this work, CS involves the collaboration of the computing discipline, the mathematical support represented by the knowledge discovery process and by the models, and the computing parallel environment capability. Central to this computational science problem is the performance prediction of data-dependent applications, as shown on Figure 1. By the way, CS allows doing things that were previously too difficult to do due to the complexity of the mathematics, the large number of calculations involved, or a combination of both. Therefore, new challenges are continuously arising although CS is still at an early stage of development.