R中的当前multiClassSummary()函数提供了没有精度的性能参数,例如如下:
示例:
predicted <- matrix(rnorm(50), ncol = 5)
observed <- rnorm(10)
apply(predicted, 2, postResample, obs = observed)
classes <- c("class1", "class2")
set.seed(1)
dat <- data.frame(obs = factor(sample(classes, 50, replace = TRUE)),
pred = factor(sample(classes, 50, replace = TRUE)),
class1 = runif(50), class2 = runif(50))
multiClassSummary(dat,lev=classes)
结果:
logLoss AUC Accuracy Kappa
0.6119378 0.4780844 0.6000000 0.1638796
F1 Sensitivity Specificity Pos_Pred_Value
0.6774194 0.7500000 0.4090909 0.6176471
Neg_Pred_Value Detection_Rate Balanced_Accuracy
0.5625000 0.4200000 0.5795455
有没有办法提供精确度?
将 Pos_Pred_Value 作为class1的精度, Neg_Pre_Value 作为class2的精度是否正确?