Confusion matrices, precision, recall, F1, ROC AUC, RMSE, calibration, imbalanced data and hyperparameter tuning.
40 cards · basic cards · AI-written, checked twice. Edit anything.
- What is a confusion matrix?
Show answer
A table showing true positives, true negatives, false positives, and false negatives for a classifier's predictions- What does a true positive (TP) represent?
Show answer
A correct prediction where the model predicted positive and the actual label is positive- What does a false positive (FP) represent?
Show answer
An incorrect prediction where the model predicted positive but the actual label is negative, also called Type I error- What does a true negative (TN) represent?
Show answer
A correct prediction where the model predicted negative and the actual label is negative- What does a false negative (FN) represent?
Show answer
An incorrect prediction where the model predicted negative but the actual label is positive, also called Type II error- What is precision?
Show answer
The proportion of positive predictions that were correct, calculated as TP / (TP + FP)- What is recall?
Show answer
The proportion of actual positives that were correctly identified, calculated as TP / (TP + FN)- What is specificity?
Show answer
The proportion of actual negatives that were correctly identified, calculated as TN / (TN + FP)- What is the F1 score?
Show answer
The harmonic mean of precision and recall, calculated as 2 * (precision * recall) / (precision + recall)- What trade-off exists between precision and recall?
Show answer
Increasing one typically decreases the other; the choice depends on the cost of false positives versus false negatives for your problem- What is a ROC curve?
Show answer
A plot of True Positive Rate (y-axis) versus False Positive Rate (x-axis) at various classification thresholds- What does AUC measure?
Show answer
The area under the ROC curve; it represents the probability that the model ranks a random positive example higher than a random negative example- What is the True Positive Rate (TPR)?
Show answer
The fraction of actual positives correctly identified, equivalent to recall, calculated as TP / (TP + FN)- What is the False Positive Rate (FPR)?
Show answer
The fraction of actual negatives incorrectly classified as positive, calculated as FP / (FP + TN)- What does an AUC of 0.5 indicate?
Show answer
The model performs no better than random chance, essentially flipping a coin
FlashKeepers