Exercise: Validation

Visual validation

• Use the Sentinel-2 data to predict the mean vegetation height for the entire Lahntal.
• Save the prediction as a tif file and have a look at it with your favorite GIS application.
• What can you say about the quality of the prediction?

Statistical validation

• Predict the mean vegetation height on your testing set.
• Create a scatterplot with the actual mean vegetation height on the x-axis and the predicted mean vegetation height on the y-axis.

One of the most used error metrics to assess the quality of a model is the Root Mean Squared Error or short the RMSE.

• How is the RMSE calculated?
• Implement the formula for the RMSE in R and calculate it for your predictions of the test data. How can you interpret this value?
• Divide the RMSE by the mean of the observed vegetation height in the test data. How can you interpret this value?