Data Archive

Confusion Matrix

Description

  • A Confusion Matrix is a table used to evaluate the performance of a Classification model. It provides a detailed breakdown of the model’s predictions across different classes, showing the number of true positives, true negatives, false positives, and false negatives.

Purpose

  • The confusion matrix helps identify where the classifier is making errors, indicating where it is “confused” in its predictions.

Structure

Structure

  • True Positives (TP): Correctly predicted positive instances.
  • False Positives (FP): Incorrectly predicted positive instances (Type 1 error).
  • True Negatives (TN): Correctly predicted negative instances.
  • False Negatives (FN): Incorrectly predicted negative instances (Type 2 error).

Metrics

  • Accuracy: The overall percentage of correct predictions. In this case, the accuracy is 78.3%.
  • Precision: The ratio of true positives to all positive predictions (including both TPs and FPs). In this case, the precision for class 0 is 85.7% and the precision for class 1 is 66.4%.
  • Recall: The ratio of true positives to all actual positive cases (including both TPs and FNs). In this case, the recall for class 0 is 80.6% and the recall for class 1 is 74.1%.
  • F1 Score: A harmonic average of precision and recall. In this case, the F1-score for class 0 is 83.0% and the F1-score for class 1 is 70.0%.
  • Specificity
  • Sensitivity

Further Examples

Example Code

from sklearn.metrics import confusion_matrix
 
# Assuming y_train and y_train_pred are your true and predicted labels
conf_matrix = confusion_matrix(y_train, y_train_pred)
print(conf_matrix)

Example Output:

array([[377, 63],
       [ 91, 180]], dtype=int64)

Graph View