Visiting Assistant Professor of Statistics Grayson White from Reed College will visit campus to discuss his latest research in improving survey estimates for the Forest Inventory and Analysis Program.
"The Forest Inventory and Analysis Program (FIA) of the US Forest Service collects survey data on the forests across the United States. In recent years, FIA has become increasingly interested in combining remotely sensed (e.g., satellite imagery and LiDAR) data with their ground-collected survey data via statistical models to help produce estimates of forest attributes in small geographic regions. In this talk, I'll first discuss typical approaches to estimation in this setting. Then, I’ll introduce a method for assessing these approaches by generating an artificial population using a weighted k-Nearest Neighbors imputation technique. Finally, based on the results of the model assessment, I will introduce a model more suitable to the data at hand: a two-stage model that accounts for model-misspecification due to zero-inflation in the data."
Join us on Monday, Oct. 26, for this exciting presentation from White. The presentation will begin in Hayes 109 at 3:10 p.m. We hope to see you there!