Data-Driven Discovery of Governing Equations for Complex Dynamical Systems by Eric Gage '24

Data-Driven Discovery of Governing Equations for Complex Dynamical Systems by Eric Gage ’24, Wednesday February 7th, 1:00 – 1:50pm, North Science Building 113, Wachenheim

In this presentation, our primary focus is to explore the capabilities of the Sparse Identification of Nonlinear Dynamical Systems (SINDy) algorithm and its corresponding Python package. Picture yourself creating dynamical systems that model phenomena such as weather patterns, transmission of diseases, or the behavior of complex mechanical systems. The SINDy algorithm is a powerful tool for identifying sparse governing equations for these systems based on empirical measurements. Through utilization of the SINDy algorithm, we can construct a model that not only facilitates predictions regarding the future behavior of a system, but also provides insight as to what is driving these systems. This process enables a comprehensive understanding of the underlying dynamics and can contribute to informed decision-making in the real-world. Not only will we look at how this process works, but we will also discuss why the sparsity aspect of the equations is crucial for the real-world application of these models. The presentation will conclude with a demonstration of numerical simulations, providing insights into how the SINDy algorithm can be employed to gain valuable information and make informed predictions in a variety of dynamical systems.