Paradigm Shift Through the Integration of Physical Methodology and Data Science
arXiv:2110.01408
Abstract
Data science methodologies, which have undergone significant developments recently, provide flexible representational performance and fast computational means to address the challenges faced by traditional scientific methodologies while revealing unprecedented challenges such as the interpretability of computations and the demand for extrapolative predictions on the amount of data. Methods that integrate traditional physical and data science methodologies are new methods of mathematical analysis that complement both methodologies and are being studied in various scientific fields. This paper highlights the significance and importance of such integrated methods from the viewpoint of scientific theory. Additionally, a comprehensive survey of specific methods and applications are conducted, and the current state of the art in relevant research fields are summarized.
arXiv comment15 pages, 7 figures
References in corpus (13)
- Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
- An Energy Approach to the Solution of Partial Differential Equations in Computational Mechanics via Machine Learning: Concepts, Implementation and Applications
- Theory-guided Data Science: A New Paradigm for Scientific Discovery from Data
- Koopman invariant subspaces and finite linear representations of nonlinear dynamical systems for control
- The Physics of Climate Variability and Climate Change
- Stable machine-learning parameterization of subgrid processes for climate modeling at a range of resolutions
- Data-Driven POD-Galerkin Reduced Order Model for Turbulent Flows
- Data-driven discovery of PDEs in complex datasets
- Hybrid modeling and prediction of dynamical systems
- Synchronization to big-data: nudging the Navier-Stokes equations for data assimilation of turbulent flows
- Lagrangian Dynamic Mode Decomposition for Construction of Reduced-Order Models of Advection-Dominated Phenomena
- Deep transfer learning in the assessment of the quality of protein models
- Hybrid Scheme of Kinematic Analysis and Lagrangian Koopman Operator Analysis for Short-term Precipitation Forecasting