Logan_et_al_2020
Logan, T. M., et al. “Night and day: The influence and relative importance of urban characteristics on remotely sensed land surface temperature.” Remote Sensing of Environment 247 (2020): 111861.
- train a random forest on spatially autocorrelated data
- used 500 trees that each tree is trained on randomly selected subset of 1/3 of the covariates - resulting prediction is unweighted mean of the prediction from all trees
- validated using holdout cross-validation (80% of data into training set, 20% into testing set)
- code provided