1. Question 1

What does LinearRegression() do?

• ❌ Assigns a model to lm

• ❌ Fits regression to lm

• ❌ Predicts output

• ✅ Creates a linear regression object (model) — but does NOT store it in a variable unless assigned

Explanation:

LinearRegression() simply creates the model object.
To store it:

---

2. Question 2

Residual plot shows errors increasing with x. This means:

• ❌ Highly accurate model

• ✅ A non-linear model may be more appropriate

• ❌ Random disruption

• ❌ Perfect regression line

Explanation:

Increasing residuals → heteroscedasticity → linear regression not suitable → try non-linear.

---

3. Question 3

PolynomialFeatures(degree=3) creates features of what order?

• ✅ Between 0 and 3 inclusive

• ❌ Maximum of 3

• ❌ Only degree 3

• ❌ Minimum of 3

Explanation:

PolynomialFeatures creates:

• degree 0 (bias)

• degree 1

• degree 2

• degree 3

---

4. Question 4

Blue regression MSE area is much smaller than red mean-line MSE area. Meaning?

• ✅ It captures data trends well.

• ❌ Adds variations

• ❌ Performs poorly

• ❌ Copies mean

Explanation:

Smaller MSE → model fits data better than using the mean.

---

5. Question 5

Romy predicts car price = $13,771.30 for 30 MPG. Next step?

• ❌ Fit model again

• ❌ Plot regression graph

• ❌ Compare residual plot

• ✅ Analyze values to validate their range

Explanation:

After prediction, verify if the predicted value is reasonable compared to dataset ranges (sanity check).

---

🧾 Summary Table

---