1. Learn
    2. /
  2. Courses
    3. /
  3. Credit Risk Modeling in R

Exercise

Creating a confusion matrix

In this example, assume that you have run a model and stored the predicted outcomes in a vector called model_pred. You want to see how the model performed so you will construct a confusion matrix. You will compare the actual loan status column (loan_status) to the predicted values (model_pred), using the table() function, where the arguments are the true values and the predicted values. Recall the confusion matrix structure:

and formulas:

$$\textrm{Classification accuracy} = \frac{(TP + TN)}{(TP + FP + TN + FN)}$$

$$\textrm{Sensitivity} = \frac{TP}{(TP + FN)}$$

$$\textrm{Specificity} = \frac{TN}{(TN + FP)}$$

Instructions

100 XP
  • Create a confusion matrix comparing the loan_status column in test_set with the vector model_pred. You can use the table() function with two arguments to do this. Store the matrix in object conf_matrix.
  • Compute the classification accuracy and print the result. You can either select the correct matrix elements from the confusion matrix using conf_matrix, or copy and paste the desired values.
  • Compute the sensitivity and print the result.