Activity Number:
|
362
- Contributed Poster Presentations: Section on Physical and Engineering Sciences
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 5, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
Abstract #313684
|
|
Title:
|
Multi-Resolution Approximations for Large Agricultural Data
|
Author(s):
|
Paul May* and Hossein Rekabdarkolaee
|
Companies:
|
and
|
Keywords:
|
Agriculture;
Spatial Statistics;
Large Data;
Regression
|
Abstract:
|
Precision agriculture is the leveraging of data for better farming practices. An important aspect of precision agriculture is analyzing the effect of covariates on crop yield. Agricultural data sets can be very large, making likelihood-based inference on traditional spatial models computationally burdensome. The Multi-Resolution Approximation allows for fast inference on Gaussian processes by using a particular covariance structure. We show through a simulation study and the analysis of a real agricultural data set that the Multi-Resolution Approximation can be used to estimate covariate effects with near-identical accuracy as traditional likelihood estimation, and with great computational advantage.
|
Authors who are presenting talks have a * after their name.