Virtual Labs

Perceptron

1. What type of decision boundary does a single-layer perceptron create?

a: A circular boundary Explanation

Explanation

b: A linear hyperplane Explanation

Explanation

c: A polynomial curve Explanation

Explanation

d: Multiple disconnected regions Explanation

Explanation

2. What does it mean for a dataset to be linearly separable?

a: The data can be separated by a straight line or hyperplane Explanation

Explanation

b: The data points are arranged in a line Explanation

Explanation

c: The dataset has only linear features Explanation

Explanation

d: All data points have the same label Explanation

Explanation

3. What is the XOR problem in the context of perceptron?

a: A linearly separable problem that perceptrons can easily solve Explanation

Explanation

b: A non-linearly separable problem that single-layer perceptrons cannot solve Explanation

Explanation

c: A problem with too much training data Explanation

Explanation

d: A convergence issue due to wrong learning rate Explanation

Explanation

4. In the perceptron learning rule, when are weights updated?

a: Only when the prediction is correct Explanation

Explanation

b: After every epoch regardless of errors Explanation

Explanation

c: Only when the perceptron makes an incorrect prediction Explanation

Explanation

d: Before making any predictions Explanation

Explanation

5. What happens to the weights when the perceptron makes a correct prediction?

a: They are reset to zero Explanation

Explanation

b: They remain unchanged Explanation

Explanation

c: They increase by the learning rate Explanation

Explanation

d: They are randomly reinitialized Explanation

Explanation

6. What role does the learning rate play in perceptron training?

a: It determines how many epochs to train Explanation

Explanation

b: It controls the step size of weight updates Explanation

Explanation

c: It sets the initial weights Explanation

Explanation

d: It defines the activation function Explanation

Explanation

7. What does the bias term 'b' allow the perceptron to do?

a: Increase the number of input features Explanation

Explanation

b: Move the decision boundary away from the origin Explanation

Explanation

c: Make the decision boundary curved Explanation

Explanation

d: Classify multiple classes simultaneously Explanation

Explanation

8. Which logic gate CAN be solved by a single-layer perceptron?

a: XOR gate Explanation

Explanation

b: XNOR gate Explanation

Explanation

c: AND gate Explanation

Explanation

d: None of the above Explanation

Explanation

9. What mathematical equation defines the decision boundary of a perceptron in 2D?

w_1 \cdot x_1 + w_2 \cdot x_2 + b = 0

Explanation

w_1 \cdot x_1^2 + w_2 \cdot x_2^2 + b = 0

Explanation

x_1 + x_2 = 1

Explanation

x_1 \times x_2 = 0

Explanation

10. What fundamental limitation did the XOR problem reveal about single-layer perceptrons?

a: They require too much training data Explanation

Explanation

b: They can only create linear decision boundaries Explanation

Explanation

c: They cannot handle binary inputs Explanation

Explanation

d: They are too slow to train Explanation

Explanation