Confusion matrix analysis in TensorFlow - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a confusion matrix in machine learning?

A confusion matrix is a table that shows how well a classification model performs by comparing actual labels with predicted labels. It helps us see where the model makes correct and wrong predictions.

beginner

What do the terms True Positive (TP), False Positive (FP), True Negative (TN), and False Negative (FN) mean in a confusion matrix?

TP: Model correctly predicts positive.<br>FP: Model wrongly predicts positive.<br>TN: Model correctly predicts negative.<br>FN: Model wrongly predicts negative.

beginner

How do you calculate accuracy from a confusion matrix?

Accuracy = (TP + TN) / (TP + TN + FP + FN). It shows the overall correctness of the model's predictions.

intermediate

What is precision and why is it important?

Precision = TP / (TP + FP). It tells us how many of the predicted positives are actually correct. Important when false positives are costly.

intermediate

How can you compute a confusion matrix using TensorFlow?

Use tf.math.confusion_matrix(labels, predictions) where labels are true classes and predictions are predicted classes. It returns a matrix showing counts of all classes, not specifically TP, FP, TN, FN.

What does a high number of False Negatives (FN) indicate?

A Model misses many positive cases

B Model predicts too many positives

C Model predicts negatives correctly

D Model has high precision

False Negatives mean the model predicted negative when the true label was positive, so it misses positive cases.

Which metric is calculated as TP / (TP + FP)?

A Recall

B Accuracy

C F1 Score

D Precision

Precision measures how many predicted positives are actually positive.

In TensorFlow, which function computes the confusion matrix?

A tf.math.confusion_matrix()

B tf.metrics.confusion_matrix()

C tf.confusion_matrix()

D tf.linalg.confusion_matrix()

tf.math.confusion_matrix() is the correct TensorFlow function to compute confusion matrices.

What does the diagonal of a confusion matrix represent?

A Wrong predictions

B False negatives

C Correct predictions

D False positives

The diagonal entries show counts where predicted labels match true labels, meaning correct predictions.

Which metric is best to use when false positives are very costly?

A Recall

B Precision

C Accuracy

D Loss

Precision focuses on reducing false positives, so it is important when false positives are costly.

Explain what a confusion matrix is and how it helps evaluate a classification model.

Think about how you check if your model guesses right or wrong.

Describe how to compute precision and recall from a confusion matrix and when each is important.

Precision is about correct positive predictions; recall is about finding all positives.