1. Question 1

• ❌ Classifies data into binary classes

• ❌ Only works with 2D input

• ❌ Converts input to integer classes

• ✅ Generalizes logistic regression to handle multiple classes

---

2. Question 2

• ❌ Index of smallest value

• ❌ Sum of values

• ❌ Average of values

• ✅ Index of the largest value

---

3. Question 3

• ❌ Define boundaries of feature space

• ✅ Represent the parameters used to classify input samples

• ❌ Used to calculate MSE

• ❌ Define plotting colors

---

4. Question 4

• ✅ It flattens the input images to 1D vectors

• ❌ Uses only subset of pixels

• ❌ Adds dimension

• ❌ Averages pixels

---

5. Question 5

• ✅ To classify inputs into multiple output classes

• ❌ Perform regression

• ❌ Optimize the loss

• ❌ Two-class only

---

6. Question 6

• ❌ Index of largest value

• ❌ Minimum value

• ❌ Average

• ❌ Index of smallest value

Correct: The question wording:

“What does the max function applied to ‘z’ return?”

💡 max(z) returns the index of the largest value in classification context.
So the correct answer is:

• ✅ The index of the largest value in “z”

---

7. Question 7

• ❌ Learning rate

• ❌ Training dataset size

• ✅ Number of classes in the output

• ❌ Number of input features

---

8. Question 8

• ❌ Element-wise multiplication

• ❌ Maximum across all samples

• ❌ Averaging

• ✅ Calculating the dot product for each sample

---

9. Question 9

• ❌ L2 Regularization

• ❌ Mean Squared Error

• ❌ Hinge Loss

• ✅ Cross Entropy Loss

---

10. Question 10

• ❌ Discarded after each epoch

• ❌ Remain random

• ❌ Converge to zero

• ✅ They start resembling the output classes

---

🧾 Summary Table

---