Learning from examples is central in science and engineering. Methods which address this problem have proven immense
power in the last decade, thanks to the availability of more data, faster computing hardware and improved algorithms. The impact is seen in numerous areas, including language processing, neuroscience, and epidemiology.
In the next decade, data-driven methods could have a similar impact on physical simulations. For instance, envision algorithms aiding the derivation of predictive and generalisable models of complex physics from data. While both physical modelling and data-driven methods are active independent research areas, relatively little attention has been paid to the intersection of the two.
The overall ambition of the DataSim project is, therefore, to develop next-generation physical simulation models that exploit modern machine learning techniques. Our aim is to achieve this goal through the development of new mathematics, algorithms and software. The secondary objective of the DataSim project is to educate a new generation of researchers working on the intersection of machine learning and scientific computing.
RCN FRINATEK Researcher Project