Date: October 4, 2024
Location: Room 3315,College of Forestry, Wildlife and Environment
Time: 13:00 – 14:00
Title: Machine Learning for Spatiotemporal Interpolation/Extrapolation
Attendees: 16
On October 4, 2024, Dr. Zutao Yang from CFWE delivered a presentation titled “Machine Learning for Spatiotemporal Interpolation/Extrapolation”. The seminar, attended by 16 people in person and via Zoom, focused on estimating the unobserved environmental data using existing datasets through interpolation and extrapolation methods.
Dr. Yang showcased the application of machine learning in spatiotemporal interpolation utilizing remote sensing and eddy covariance flux data. He emphasized how these techniques can model complex patterns, manage non-linear relationships, and effectively leverage large datasets, making a significant improvement compared to traditional methods. Through several case studies, Dr. Yang illustrated how machine learning methods can significantly enhance the accuracy and robustness of interpolation and extrapolation, offering valuable insights into spatiotemporal analysis.
The seminar concluded with a Q & A session, where faculty and students asked several questions such as: What caused data gaps? What is the distance between monitoring equipment for other variables? Have you used other data for gap filling? What is the maximum observed gap in the data?
We thank Dr. Yang for his engaging presentation and look forward to more discussions, research efforts, and collaborations in AI, ML, and related fields!