|
Activity Number:
|
420
- Modern Modeling Approaches for Imputation Using Survey Data
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 12, 2021 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #316575
|
|
|
Title:
|
Design Consistent Random Forest Models for Survey Data
|
|
Author(s):
|
daniell toth* and Kelly McConville
|
|
Companies:
|
US Bureau of Labor Statistics and Reed College
|
|
Keywords:
|
desgin consistent;
sample design;
tree models;
machine learning;
small area;
official statistics
|
|
Abstract:
|
Random forest models represent a useful and flexible tool for producing a nonparametic model that can provide accurately predicted values. There are many potential applications for these types of models when dealing with survey data. However, survey data is usually collected using a complex sample design, so it is necessary to have an algorithm for creating random forest models that account for this sample design during model estimation. In this article, we provided an algorithm to produce consistent forest models under complex sample designs and explore their use for producing small-domain estimates for official statistics.
|
Authors who are presenting talks have a * after their name.
