This tech talk has been canceled and will be rescheduled. We apologize for the inconvenience!
Abstract:
Computational science uses a computer’s super power and mathematical algorithms to solve large-scale scientific problems. Data science explores information from large quantity heterogeneous datasets to gain insights and build forecast models with statistical methods. Wouldn’t it be great to combine the strengths from both worlds?
In this talk, I will provide a brief introduction of computational science and data science, then show a selection of the research projects I was involved in on the subjects of computational and data science applied to air quality modeling, presidential elections, predictive policing, and DNA binding hotspot forecasting. In all of these projects, mathematics and computer science play important roles, together with problem-solving skills and subject knowledge from various disciplines. Inspired by those projects, I developed and taught data science classes for both undergraduate and MBA programs at Willamette University. In those classes, teams of students worked on various projects, which demonstrated that computational and data science projects are intrinsically collaborative, and are indeed relevant to all.
Bio:
Haiyan Cheng is a Computer Science professor and department chair at Willamette University. She received her Master’s degree in Applied Mathematics from Michigan Technological University, a second Master’s degree in Computer Science from the University of Windsor, Canada, and a Ph.D. in Computer Science and Applications from Virginia Tech. Her research interests are in uncertainty quantification and reduction techniques for large scale simulations. She has applied the Polynomial Chaos method to Sulfur Transport Eulerian Model (STEM) and proposed hybrid 4D-Variational and Ensemble Kalman Filter (EnKF) model for data assimilations. Her research on nonlinear particle filters was supported by a grant from the National Science Foundation. She has also worked on trust delegation using the Dempster-Shafer theory for Open Grid Service Architecture (OGSA) single-sign-on. In recent years, she has participated in various research projects in applied data science, as well as taught data science classes. Currently, she is exploring techniques that effectively combine model-driven and data-driven approaches.
She has publications in “Research in Data Science”, “The Journal of Computing Sciences in College”, “Atmosphere”, “International Journal for Uncertainty Quantification”, “Environmental Modeling and Software”, “Tellus Series A: Dynamic Meteorology and Oceanography”, “Mathematics and Computers in Simulation”, and “Integral Methods in Science and Engineering”.
This talk will be live-streamed at: https://www.youtube.com/c/GaloisInc/live