Skip to main content

Table 2 Performance of the four classification methods used to predict the prevalence of B. afzelii, B. garinii, and B. valaisiana in a grid of 0.25° radius over Europe using a combination of environmental data

From: An updated meta-analysis of the distribution and prevalence of Borrelia burgdorferi s.l. in ticks in Europe

  B. afzelii B. garinii B. valaisiana
  AUC CA AUC CA AUC CA
Neural Networks 0.491 0.972 0.585 0.963 0.496 0.989
Naive Bayes 0.550 0.626 0.592 0.662 0.824 0.775
AdaBoost 0.499 0.974 0.500 0.961 0.500 0.989
SVM 0.816 0.987 0.816 0.981 0.844 0.995
  1. AUC area under the curve, CA classification accuracy. The AUC values refer to the average of all the classifications. Complete AUC and CA values for each category can be obtained from the Additional files 2 and 3