Activity Number:
|
248
- Machine Learning in Science and Industry
|
Type:
|
Contributed
|
Date/Time:
|
Monday, July 29, 2019 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
Abstract #304636
|
Presentation
|
Title:
|
A Machine-Learning Approach to Extract Remote-Sensing Features for Predicting Crop Yield
|
Author(s):
|
Luca Sartore* and Arthur Rosales and David Johnson and Mary Frances Dorn and Clifford Spiegelman
|
Companies:
|
National Institute of Statistical Sciences and National Agricultural Statistics Service and National Agricultural Statistics Service and Los Alamos National Laboratory and Texas A&M University
|
Keywords:
|
Satellite data;
Non-parametric models;
Crop yield prediction;
Kullback-Leibler distance;
Features extraction;
County-level estimates
|
Abstract:
|
The USDA's National Agricultural Statistics Service (NASS) produces annual yield estimates for major crops at national, state, agricultural district, and county levels. Several surveys are conducted to produce reliable estimates that are further enhanced by incorporating remote sensing measurements at the county level. These measurements consist of Moderate Resolution Imaging Spectroradiometer (MODIS) data in the form of Normalized Difference Vegetative Index (NDVI) calculated from multispectral composites and Land Surface Temperature (LST) based on thermal composites. Both NDVI and LST are summarized for crop regions throughout the growing season. Corn-yield predictions at the county level are then produced using non-parametric models that combine spatial coordinates with satellite data. Random forests and boosted trees are compared when processing additional information based on Kullback-Leibler distances. A method to compute standard errors is also described, and the accuracy of these models is evaluated using established validation techniques.
|
Authors who are presenting talks have a * after their name.