Skip to main content

Table 2 Classification performance of RF model for the top 25 variables. Confusion matrix for the RF model trained with the top 25-predictor variables selected after variable importance analysis, estimated by averaging the results over ten repetitions of 10-fold cross-validation. Standard deviations are given in parenthesis *

From: What variables are important in predicting bovine viral diarrhea virus? A random forest approach

  

Real

  

BVDV-negative

BVDV-positive

Predicted

BVDV-negative

114.55 (6.8)

2.8 (0.20)

BVDV-positive

53.95 (6.8)

4.7 (0.20)

  1. *Performance metrics: ACC: 67.75 (Sd. 3.69); SPE: 67.98 (Sd. 3.85) and SEN: 62.26 (Sd. 3.33).