<Li> False negative = incorrectly rejected </Li> <P> Let us consider a group with P positive instances and N negative instances of some condition . The four outcomes can be formulated in a 2 × 2 contingency table or confusion matrix, as follows: </P> <Table> <Tr> <Td_colspan="2"> </Td> <Td_colspan="2"> True condition </Td> </Tr> <Tr> <Td> </Td> <Td> Total population </Td> <Td> Condition positive </Td> <Td> Condition negative </Td> <Td> Prevalence = Σ Condition positive / Σ Total population </Td> <Td_colspan="2"> Accuracy (ACC) = Σ True positive + Σ True negative / Σ Total population </Td> </Tr> <Tr> <Td> Predicted condition </Td> <Td> Predicted condition positive </Td> <Td> True positive </Td> <Td> False positive, Type I error </Td> <Td> Positive predictive value (PPV), Precision = Σ True positive / Σ Predicted condition positive </Td> <Td_colspan="2"> False discovery rate (FDR), probability of false alarm = Σ False positive / Σ Predicted condition positive </Td> </Tr> <Tr> <Td> Predicted condition negative </Td> <Td> False negative, Type II error </Td> <Td> True negative </Td> <Td> False omission rate (FOR) = Σ False negative / Σ Predicted condition negative </Td> <Td_colspan="2"> Negative predictive value (NPV) = Σ True negative / Σ Predicted condition negative </Td> </Tr> <Tr> <Td_colspan="2"> Click thumbnail for interactive chart: </Td> <Td> True positive rate (TPR), Recall, Sensitivity, probability of detection = Σ True positive / Σ Condition positive </Td> <Td> False positive rate (FPR), Fall - out = Σ False positive / Σ Condition negative </Td> <Td> Positive likelihood ratio (LR+) = TPR / FPR </Td> <Td> Diagnostic odds ratio (DOR) = LR+ / LR − </Td> <Td> F score = 2 / 1 / Recall + 1 / Precision </Td> </Tr> <Tr> <Td> False negative rate (FNR), Miss rate = Σ False negative / Σ Condition positive </Td> <Td> True negative rate (TNR), Specificity (SPC) = Σ True negative / Σ Condition negative </Td> <Td> Negative likelihood ratio (LR −) = FNR / TNR </Td> </Tr> </Table> <Tr> <Td_colspan="2"> </Td> <Td_colspan="2"> True condition </Td> </Tr>

Test for validity using the rules method some m are p all s are m some s are p