作者: Elizabeth A. Freeman , Gretchen Moisen
关键词: Kappa 、 Diagnostic test 、 Sensitivity (control systems) 、 Plot (graphics) 、 Measure (mathematics) 、 Algorithm 、 Binary classification 、 Set (abstract data type) 、 Receiver operating characteristic 、 Statistics 、 Mathematics
摘要: The PresenceAbsence package for R provides a set of functions useful when evaluating the results presence-absence analysis, example, models species distribution or analysis diagnostic tests. toolkit selecting optimal threshold translating probability surface into maps specifically tailored to their intended use. includes calculating dependent measures such as confusion matrices, percent correctly classified (PCC), sensitivity, specificity, and Kappa, produces plots each measure is varied. It also plot Receiver Operator Characteristic (ROC) curve calculates associated area under (AUC), independent model quality. Finally, computes thresholds by multiple criteria, these optimized on graphs.