Challenge - 5 Problems

🎖️

ROC and AUC Mastery

Get all challenges correct to earn this badge!

Test your skills under time pressure!

🧠 Conceptual

intermediate

2:00remaining

Understanding the ROC Curve

What does the ROC curve represent in a binary classification model?

AThe trade-off between true positive rate and false positive rate at various thresholds

BThe relationship between precision and recall for different classes

CThe distribution of predicted probabilities for positive class only

DThe accuracy of the model over different training epochs

Attempts:

2 left

❓ Predict Output

intermediate

2:00remaining

Calculating AUC from Model Predictions

Given the following true labels and predicted probabilities, what is the AUC score?

y_true = [0, 0, 1, 1]
y_scores = [0.1, 0.4, 0.35, 0.8]

ML Python

from sklearn.metrics import roc_auc_score

y_true = [0, 0, 1, 1]
y_scores = [0.1, 0.4, 0.35, 0.8]
auc = roc_auc_score(y_true, y_scores)
print(round(auc, 2))

A0.95

B0.85

C0.65

D0.75

Attempts:

2 left

❓ Metrics

advanced

2:00remaining

Interpreting AUC Values

Which statement best describes an AUC value of 0.5 for a binary classifier?

AThe model perfectly separates positive and negative classes

BThe model performs no better than random guessing

CThe model has high precision but low recall

DThe model has a high false positive rate

Attempts:

2 left

🔧 Debug

advanced

2:00remaining

Fixing Incorrect ROC Curve Plot

Consider this code snippet to plot an ROC curve. What is the main issue causing the plot to be incorrect?

from sklearn.metrics import roc_curve
import matplotlib.pyplot as plt

y_true = [0, 1, 0, 1]
y_scores = [0.2, 0.8, 0.4, 0.6]
fpr, tpr, thresholds = roc_curve(y_true, y_true)
plt.plot(fpr, tpr)
plt.show()

AThe roc_curve function is called with true labels instead of predicted scores

BThe matplotlib import is incorrect

CThe y_true list contains invalid values

DThe plot function is missing labels for axes

Attempts:

2 left

❓ Model Choice

expert

3:00remaining

Choosing Model Based on AUC for Imbalanced Data

You have two binary classifiers evaluated on a highly imbalanced dataset. Model A has an AUC of 0.92, and Model B has an AUC of 0.85. However, Model B has higher precision on the minority class. Which model should you choose if your priority is to correctly identify positive cases while minimizing false alarms?

AChoose Model A because it has higher recall on the minority class

BChoose Model A because higher AUC means better overall ranking

CChoose Model B because higher precision reduces false alarms on positives

DChoose Model B because AUC is not useful for imbalanced data

Attempts:

2 left