TABLE 4.

Validation metrics of the models under study on the withheld gull dataset.

Method	Accuracy	Precision	Recall	Variable selection probability
Variable/Covariate	0.97	0.1	0.99	0.71
Constant	0.97	0.1	0.99	–
Intercept	0.97	0.1	0.99	–
Main	0.97	0.1	0.99	0.29
Fixed intercov	0.97	0.1	0.99	0.70
Fixed covariate	0.97	0.1	0.99	0.71
Machine Learning	0.89	0.8	0.90	–

Note: The accuracy is the proportion of correctly classified validated data, the precision is the proportion of mismatched identities that were correctly matched and recall is the proportion of correctly matched identities that were recovered. The number of validated samples was 384 out of which 10 were mismatched.