Random Forest Spatial Interpolation

Authors: David Lamb*, University of South Florida
Topics: Geographic Information Science and Systems, Spatial Analysis & Modeling
Keywords: Random Forest, Decision Tree, Interpolation
Session Type: Paper
Day: 4/11/2018
Start / End Time: 3:20 PM / 5:00 PM
Room: Riverview II, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded


Spatial interpolation are powerful methods for deriving continuous data from discrete points. While there are a variety of approaches, few allow for high-dimensional data to be used in predicting the interpolated value. Recent studies have shown decision tree regression to be a potential method for predicting values across a region. Ensemble machine learning methods like random forests allow for an optimal decision tree to be selected from a set. Parameters may also be identified that minimize a scoring function (such as mean square error). This presentation will provide a background on using a Random Forest Regression to create interpolated surfaces, how additional variables can be included in the regressor, and potential best practices.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login