Table 4.

Prediction accuracy of supervised models. Accuracy is evaluated by the average of a ten-fold cross validation, for various combinations of training/test sets

Model	Can/Can	Cali/Cali	Can/Cali	Cali/Can	All/Can	All/Cali
Logistic regression	0.927	0.883	0.811	0.914	0.915	0.811
Random forest classifier	0.896	0.860	0.816	0.898	0.909	0.802
K-Neighbors classifier	0.852	0.827	0.788	0.884	0.885	0.788
SVC	0.904	0.813	0.805	0.909	0.909	0.806
Gaussian process classifier	0.919	0.810	0.792	0.911	0.911	0.792
AdaBoost classifier	0.860	0.770	0.810	0.911	0.911	0.811
XGB classifier	0.880	0.847	0.778	0.904	0.904	0.778

Note: Can for Canada, Cali for California, and All for the combined dataset