What is the difference between a confusion matrix and a classification report?

2 Classification report

A classification report is a text summary that shows the main metrics for each class of a machine learning model. It usually includes the precision, recall, F1-score, and support for each class, as well as the weighted average of these metrics across all classes. For example, if you have the same binary classification model as before, a classification report might look like this:

|              | Precision | Recall | F1-score | Support |
|--------------|-----------|--------|----------|---------|
| Spam         | 0.89      | 0.80   | 0.84     | 100     |
| Not Spam     | 0.82      | 0.90   | 0.86     | 100     |
|--------------|-----------|--------|----------|---------|
| Weighted Avg | 0.85      | 0.85   | 0.85     | 200     |        

A classification report gives you a detailed breakdown of how well your model performs on each class, and how it balances the trade-off between precision and recall. It also shows you the number of instances (support) for each class, which can indicate the class imbalance or the size of the dataset.

4 When to use confusion matrix and classification report

Confusion matrix and classification report are both useful tools for machine learning evaluation and validation, but they serve different purposes for different audiences. When wanting to explore the distribution and frequency of the predictions and outcomes, detect misclassification between classes, compare the performance of different models or algorithms, or visualize the results with heatmaps, bar charts, or pie charts, a confusion matrix is ideal. On the other hand, a classification report is more suitable when measuring the precision, recall, F1-score, and support of each class; assessing the trade-off between precision and recall; evaluating the weighted average of metrics across all classes; or reporting results with tables, numbers, or text.

5 How to generate confusion matrix and classification report

There are many ways to generate a confusion matrix and a classification report, depending on the programming language, library, or tool you use. For example, if you use Python, you can use the sklearn.metrics module to import the confusion_matrix and classification_report functions. Then, you can pass the actual and predicted values of your model as arguments to these functions, and print or plot the results. Here is a simple code snippet that shows how to do that:

# Import the functions
from sklearn.metrics import confusion_matrix, classification_report
# Define the actual and predicted values
y_true = [0, 0, 1, 1, 0, 1, 0, 1, 1, 0]
y_pred = [0, 1, 1, 0, 0, 1, 0, 1, 1, 1]
# Generate and print the confusion matrix
cm = confusion_matrix(y_true, y_pred)
print(cm)
# Generate and print the classification report
cr = classification_report(y_true, y_pred)
print(cr)        

6 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?