R
決定木 Decision Tree Analysis
caret
!!!confusionMatrix()
* caretパッケージに含まれる
* 混同行列（予測値と観測値の対応表）をもとに分類の精度を出す
* C2, C3, JPの分類の例
** モデル <- ctree(観測値 ~ 説明変数)

{{pre
CM.table <- table(predict(モデル), データ$観測値)

CM.table

     C2 C3 JP
  C2 35  7  0
  C3  3 19  1
  JP  0  6 34

CM.result <- caret::confusionMatrix(CM.table)

print(CM.result)

Confusion Matrix and Statistics

    
     C2 C3 JP
  C2 35  7  0
  C3  3 19  1
  JP  0  6 34

Overall Statistics
                                          
               Accuracy : 0.8381          
                 95% CI : (0.7535, 0.9028)
    No Information Rate : 0.3619          
    P-Value [Acc > NIR] : < 2.2e-16       
                                          
                  Kappa : 0.7552          
                                          
 Mcnemar's Test P-Value : NA              

Statistics by Class:

                     Class: C2 Class: C3 Class: JP
Sensitivity             0.9211    0.5938    0.9714
Specificity             0.8955    0.9452    0.9143
Pos Pred Value          0.8333    0.8261    0.8500
Neg Pred Value          0.9524    0.8415    0.9846
Prevalence              0.3619    0.3048    0.3333
Detection Rate          0.3333    0.1810    0.3238
Detection Prevalence    0.4000    0.2190    0.3810
Balanced Accuracy       0.9083    0.7695    0.9429
}}