Microsoft Azure AI Fundamentals (AI-900)Describe fundamental principles of machine learning on AzureHard
A data scientist is evaluating various machine learning models for a binary classification task. To thoroughly assess the models' performance across different classification thresholds and understand their ability to distinguish between positive and negative classes, the data scientist plots the True Positive Rate against the False Positive Rate. Which evaluation curve is the data scientist using?
- APrecision-Recall Curve
- BROC Curve
- CLift Curve
- DGain Curve
Show answer & explanationAnswer & explanation
Correct answer: B. ROC Curve
The ROC (Receiver Operating Characteristic) curve plots the True Positive Rate (Sensitivity) against the False Positive Rate (1-Specificity) at various threshold settings. It is used to evaluate the performance of a binary classifier system as its discrimination threshold is varied.
Why the other options are wrong
- A. The Precision-Recall Curve plots Precision against Recall, which is particularly useful for imbalanced datasets, but not TPR vs FPR.
- C. Lift Curve measures how much more likely we are to receive a positive response than if we did not use the model, not TPR vs FPR.
- D. Gain Curve (or Cumulative Gains Curve) shows the proportion of targets identified when considering a given percentage of the population, not TPR vs FPR.
ROC Curve (Receiver Operating Characteristic)
A graph showing the performance of a classification model at all classification thresholds. It plots two parameters: True Positive Rate (TPR) on the y-axis and False Positive Rate (FPR) on the x-axis.
- TPR is also known as Sensitivity or Recall.
- FPR is calculated as 1 - Specificity.
- The Area Under the Curve (AUC-ROC) provides a single metric for overall performance.
- Useful for evaluating models on both balanced and imbalanced datasets.
Memory trick: Drawing lines to see how good our predictions are.