Twin Cities WRS Seminar
Integrating Physical Knowledge into Machine Learning: Applications in Lake Temperature Prediction
Graduate Student, Computer Science and Engineering
February 7, 2020
Physics-based models have been widely used to study engineering and environmental systems in domains such as hydrology, climate science, materials science, agriculture, and computational chemistry. Despite their extensive use, these models have several well-known limitations due to simplified representations of the physical processes being modeled or challenges in selecting appropriate parameters. Given rapid data growth due to advances in sensor technologies, there is a tremendous opportunity to systematically advance modeling in these domains by using machine learning (ML) methods. However, capturing this opportunity is contingent on a paradigm shift in data-intensive scientific discovery since the "black box" use of ML often leads to serious false discoveries in scientific applications. This work presents a novel methodology for combining physics-based models with state-of-the-art deep learning methods to leverage their complementary strengths. Moreover, we show that meta-learning frameworks in machine learning can be leveraged to accurately predict environment variables of unmonitored systems using models from data-rich systems and contextual data.