Hands-On Exercise: Implement a Simple Perceptron CT-AI Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A developer is implementing a perceptron with 3 inputs. The current weights are [0.2, -0.4, 0.6] and the bias is 0.1. The input vector is [1, 2, 3]. What is the weighted sum before applying the activation function?

   • A0.5
   • B1.2
   • C1.3
   • D2.4
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   In the mathematical formulation of a simple perceptron, what is the role of the bias term?

   • AIt shifts the decision boundary away from the origin, allowing the perceptron to classify data that is not centered at zero.
   • BIt multiplies the activation function output to produce the final prediction.
   • CIt determines the learning rate used to update the weights during training.
   • DIt scales the input features to a common range before they are weighted.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A QA engineer is implementing a perceptron to classify emails as spam (1) or not spam (0) based on two features: word count and number of links. The engineer initializes weights to 0 and bias to 0. The first training sample has features [3, 1], true label 1, and the activation function is a step function that outputs 1 if the weighted sum is greater than 0, otherwise 0. What is the predicted output for this first sample before any weight update?

   • A3
   • B0
   • C1
   • D4
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   What is the purpose of the learning rate in the perceptron learning rule?

   • AIt controls how much the weights are adjusted in response to each prediction error.
   • BIt determines the number of training epochs required for convergence.
   • CIt decides whether the perceptron uses batch or stochastic gradient descent.
   • DIt sets the threshold for the activation function.
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A test engineer is writing a test suite for a perceptron implementation. The test suite includes a test that feeds a single sample with features [1, 1] and true label 1, and asserts that the weights are updated. However, the test fails intermittently. What is the most likely reason?

   • AThe learning rate is set to 0, so no weight update occurs.
   • BThe test data is not linearly separable, so the perceptron cannot learn.
   • CThe activation function is incorrectly implemented as a linear function.
   • DThe test does not reset the weights and bias to a known initial state before each test run.
   Reveal Solution Discussion Select an answer first