Course module: Regression modelling for digital soil mapping

Course title: Regression modelling for digital soil mapping
Course Code:  
Subject areas (Curriculum): Soil and environmental sciences
Duration: 1 day
Short description: This module introduces regression modelling for digital soil mapping, focussing on linear regression, CART and random forests modelling.
Moderators: Bas Kempen
Target audience: Soil and environmental scientists with an interest in quantitative methods for mapping soil and environmental properties.
Requirements: Basic understanding of statistics; basic knowledge of the statistical software R
Programme: Lecture (2h); Computer exercises (6h)
Objective: Introducing regression modelling for digital soil mapping: model theory and assumptions, model selection, random forest modelling
Software / materials:

R, RStudio

R packages: gstat, sp, randomForest, plyr, e1071, ggplot2, aqp, rgdal, GSIF, plotKML

Case studies: Edgeroi (Australia)
Literature:  
Minimum number of participants: 10
Maximum number of participants: 35

 

Contact: