| Fact\Predict | Class A | Class B |
| Class A | True Positive | False Negative |
| Class B | False Positive | True Nagative |
A confusion table for Class A
Positive/ Negative: if target class is A, then the predict A is Positve, Others are negative.
True (P/N): if Predict = Fact, then it‘s True.
a. Accuracy = TN+TP/ALL
comments: not good measure when data are unbalanced.
b. True Positive Rate/ recall/ sensitivity = TP / TP + FN
comments: use it when Positive results are important
c. True Negative Rate = TN / TN + FP
R for Confusion Matrix:
library(SDMTools)
confusion.matrix(svmmodel.truth,svmmodel.class)
y: sensitivity
x: specificity
The bigger the Area of ROC is, the more accurate the model is.
原文：http://www.cnblogs.com/fuxiaotong/p/7189975.html