Table 1 Classification metrics on test data. The standard deviation was computed based on a sample size of N = 10.
From: Case study on climate change effects and food security in Southeast Asia
Model | Classification threshold | Balanced accuracy | Precision | Recall | ROC-AUC |
|---|---|---|---|---|---|
Cropland suitability via climate conditions | |||||
Logistic Regression44 | 0.15 | 0.750 | 0.296 | 0.610 | 0.871 |
Random Forest Classifier45 | 0.24 | 0.699 | 0.657 | 0.415 | 0.914 |
Naive Bayes46 | 0.98 | 0.733 | 0.376 | 0.534 | 0.891 |
MLP Classifier47 | 0.24 | 0.797 | 0.455 | 0.651 | 0.909 |
AdaBoost Classifier48 | 0.32 | 0.770 | 0.437 | 0.598 | 0.714 |
CatBoost Classifier49 | 0.32 | 0.802 | 0.660 | 0.624 | 0.959 |
XGBClassifier42 | 0.34 | 0.813 | 0.651 | 0.653 | 0.960 |
Convolutional Neural Network50 | 0.31 | 0.792 | 0.669 | 0.501 | 0.822 |
Cropland suitability via climate and socioeconomic conditions | |||||
XGBClassifier | \(0.49 \pm 1.8{\times }10^{-2}\) | \(0.959 \pm 1.3{\times }10^{-6}\) | \(0.939 \pm 3.2{\times }10^{-6}\) | \(0.924 \pm 2.8{\times }10^{-6}\) | \(0.992 \pm 5.3{\times }10^{-5}\) |
