Link Search Menu Expand Document

Here we are going to see calculations of Error Metrics in python.
Import confusion matrix from sklearn library.

from sklearn.metrics import confusion_matrix
CM1 = confusion_matrix(y_test,predictions) // passing arguements (predictions and actual values)

Build confusion matrix using crosstab function.
CM1 = pd.crosstab(y_test, predictions)

Now let us save True Positive, True Negative, False Positive, False Negative

TP = CM2.iloc[1:1]
TN = CM2.iloc[0:0]
FP = CM2.iloc[0:1]
FN = CM2.iloc[1:0]

Here

TP = 4
TN = 4
FP = 0
FN = 0

Now let us calculate classification error metrics.

Classification Metrics

  • Confusion Matrix
  • Accuracy
  • Misclassification Error
  • Specificity

  • Recall

  • we built confusion matrix.
  • Accuracy = TP+ TN/ Total Observations (TP+TN+FP+FN) = ((4+4) / 8 )*100 = 100% here.
  • Misclassification Error = FP+FN / Total = 0+0 / 8 = 0 here.
  • Specificity = TN/TN+FP = (4/4+0 )*100= 100% here.
  • Recall = TP/TP+FN = (4/4+0) *100 = 100% here.
  • False Negative Rate = FN100/FN +TP = 0100 / 0+4 = 0% here.

See below screenshot for python implementation.