Skip to main content

Table 4 Confusion matrix for Random-From-Train verbal autopsy.

From: Measuring causes of death in populations: a new metric that corrects cause-specific mortality fractions for chance

a) Random-From-Train Confusion Matrix
   Predicted   
   Stroke Diabetes Other
True Stroke 28 14 224
  Diabetes 11 13 122
  Other 223 106 1961
  1. The confusion matrix for Random-From-Train (with a length-three cause-list for clarity). As in Table 2, the entry in each cell counts the number of deaths truly due to the row cause that were predicted to be due to the column cause. This table demonstrates that while Random-From-Train is inaccurate at the individual level, at the population level the prediction distribution can closely match the truth