1️⃣ Question 1

In medical diagnosis, avoiding missed true positives is critical. Which metric matters most?

• ✅ Recall

• ❌ Accuracy

• ❌ F1 Score

• ❌ Precision

Explanation:
Recall measures how many actual positive cases you successfully identify, making it essential when missing a true positive is dangerous.

---

2️⃣ Question 2

Which metric is the square root of MSE?

• ❌ R-squared

• ❌ MAE

• ❌ MSE

• ✅ Root Mean Squared Error (RMSE)

Explanation:
RMSE = √MSE.

---

3️⃣ Question 3

Best metric to evaluate cluster separation?

• ✅ Silhouette score

• ❌ Elbow method

• ❌ Evaluation method

• ❌ Davies-Bouldin Index

Explanation:
Silhouette score measures intra-cluster similarity vs. inter-cluster separation.

---

4️⃣ Question 4

Model performs well on training but poorly on test data.

• ✅ Overfitting

• ❌ Cross-validation

• ❌ Train-test split

• ❌ Data snooping

Explanation:
Overfitting = memorizing training data instead of learning general patterns.

---

5️⃣ Question 5

Difference between Lasso and Ridge?

• ❌ Lasso only for feature selection

• ✅ Lasso = L1 penalty, Ridge = L2 penalty

• ❌ Lasso uses larger datasets

• ❌ Ridge = L1, Lasso = L2

Explanation:
L1 (Lasso) can shrink coefficients to zero → feature selection.
L2 (Ridge) shrinks coefficients but never to zero.

---

6️⃣ Question 6

How to mitigate data leakage?

• ❌ Use one feature only

• ✅ Avoid including features derived from the entire dataset

• ❌ Shuffle data

• ❌ Only separate train/test sets

Explanation:
Features like global averages or information from future data leak test information into training.

---

7️⃣ Question 7

Interpreting feature importance without considering relationships can cause:

• ❌ Scaling issues

• ❌ Minimal target variables

• ❌ Use only uncorrelated features

• ✅ Overlooking correlated features in importance scores

Explanation:
Correlated features can fool feature importance scores (importance split between them).

---

🧾 Summary Table