Faculty Mentor(s)
Dr. Guangliang Chen, Mathematics & Statistics
Document Type
Poster
Event Date
4-17-2026
Abstract
This project focuses on identifying patterns of seasonal transitions and nutrient salt fluctuations within the watershed environment. To capture the complex relationships between multiple sampling sites and environmental variables, we represent the watershed dataset as a weighted graph, where nodes correspond to water samples and edge weights reflect similarity in environmental conditions or nutrient concentrations. Using the Gaussian kernel function, we encode the connectivity structure of this network and quantify how similar different nodes are. We then perform spectral embedding by projecting the high-dimensional graph into a lower-dimensional space using the eigenvectors of the Laplacian matrix. This approach preserves the original proximity of the data while allowing us to visualize and analyze seasonal transitions more intuitively. It also provides a data-driven framework for uncovering seasonal and nutrient dynamics in watershed systems.
Recommended Citation
Repository citation: Luo, Yifan, "Analyzing Label Structure and Regional Similarity in Watershed Data via Spectral Clustering" (2026). 25th Annual A. Paul and Carol C. Schaap Celebration of Undergraduate Research and Creative Activity (2026). Paper 27.
https://digitalcommons.hope.edu/curca_25/27
April 17, 2026. Copyright © 2026 Hope College, Holland, Michigan.
Comments
This research was supported by the Jay Folkert & Charles Steketee Mathematics Summer Research Fund, Hope College Math Department and Global Water Research Institute.
Title on poster differs from abstract booklet. Poster title: Watershed data analysis by using language modeling techniques