Data-driven scientific and engineering computing algorithms and software

This research project is committed to exploring and innovating the applications of data science, artificial intelligence, and their interdisciplinary use in scientific and engineering computing algorithms and software. Our focus is on designing an innovative network structure to build surrogate models of parameterized partial differential equation solutions, thereby significantly enhancing the efficiency of scientific and engineering computations. We plan to develop an optimal experimental design strategy based on neural networks, continuously optimizing the experimental design to improve model accuracy, further promoting the application of data in scientific and engineering computations. In order to effectively handle high-dimensional data, we plan to combine deep learning technologies and traditional mathematical methods, studying dimensionality reduction techniques within a Bayesian framework, and developing new high-dimensional sampling methods. We aim to innovate the operational environment and construct a brand-new scientific computing software framework. Through this project, we look forward to promoting the application of data science and artificial intelligence in the research of scientific and engineering computing software, providing new perspectives and tools for the advancement of scientific research and industry.