Skip to main content

Advertisement

Table 6 Mean scores using WHO categories on the combined MDS and RCT datasets for each of the four classifiers

From: Automatically determining cause of death from verbal autopsy narratives

  Precision Sensitivity F1 PCCC CSMFA CCCSMFA
Adult (15–69 years)
Naïve Bayes .591 .593 .580 .583 .869 .643
Random forest .644 .647 .634 .638 .905 .742
SVM .665 .662 .655 .654 .908 .751
Neural network .630 .654 .620 .646 .840 .567
Child (29 days–14 years)
Naïve Bayes .493 .402 .427 .379 .768 .369
Random forest .570 .507 .514 .488 .807 .476
SVM .567 .530 .528 .512 .796 .446
Neural network .512 .494 .474 .474 .753 .330
Neonate (<29 days)
Naïve Bayes .434 .469 .435 .399 .797 .448
Random forest .424 .455 .426 .384 .798 .450
SVM .505 .497 .476 .431 .813 .492
Neural network .328 .361 .306 .278 .634 .007
  1. Adult: 48 categories, child: 39 categories, neonate: 17 categories. Bold indicates the best score in each column for each age group. PCCC: partially chance-corrected concordance, CSMFA: cause-specific mortality fraction (CSMF) accuracy, CCCSMFA: chance-corrected CSMFA