Forward propagation in ML Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of a simple forward propagation step

Given the following code for a single neuron forward propagation, what is the output value?

ML Python

import numpy as np

inputs = np.array([1.0, 2.0, 3.0])
weights = np.array([0.2, 0.5, -0.3])
bias = 0.1

output = np.dot(inputs, weights) + bias
print(output)

A 1.8

B 1.9

C 0.4

D 2.0

💡 Hint

Remember to multiply each input by its weight, sum them, then add the bias.

The dot product is (1.0*0.2) + (2.0*0.5) + (3.0*(-0.3)) = 0.2 + 1.0 - 0.9 = 0.3. Adding bias 0.1 gives 0.4.

❓ Model Choice

intermediate

Choosing the correct activation function for forward propagation

Which activation function is best suited for a binary classification problem during forward propagation?

A Softmax

B ReLU (Rectified Linear Unit)

C Linear

D Sigmoid

💡 Hint

Think about output values between 0 and 1 representing probabilities.

Sigmoid outputs values between 0 and 1, which can be interpreted as probabilities for binary classification.

❓ Hyperparameter

advanced

Effect of learning rate on forward propagation outputs

If the learning rate is set too high during training, how does it affect the forward propagation outputs in subsequent iterations?

A Outputs oscillate and may not converge

B Outputs remain unchanged regardless of learning rate

C Outputs become more stable and converge faster

D Outputs always decrease in value

💡 Hint

Consider how big steps in weight updates affect predictions.

A high learning rate causes large weight updates, making outputs jump around and preventing convergence.

❓ Metrics

advanced

Calculating accuracy from forward propagation predictions

A model outputs the following predictions after forward propagation for 5 samples: [0.9, 0.4, 0.6, 0.3, 0.8]. The true labels are [1, 0, 1, 0, 1]. Using a threshold of 0.5, what is the accuracy?

A 40%

B 100%

C 60%

D 80%

💡 Hint

Convert predictions to 0 or 1 using threshold, then compare to true labels.

Predictions thresholded at 0.5: [1, 0, 1, 0, 1], which exactly matches the true labels [1, 0, 1, 0, 1]. So 5 out of 5 correct = 100%.

🔧 Debug

expert

Identifying the bug in forward propagation code

What error does the following code raise during forward propagation?

ML Python

import numpy as np

def forward_prop(inputs, weights, bias):
    return np.dot(weights, inputs) + bias

inputs = np.array([1, 2, 3])
weights = np.array([0.2, 0.5])
bias = 0.1
output = forward_prop(inputs, weights, bias)
print(output)

A ValueError: shapes (2,) and (3,) not aligned for dot product

B NameError: name 'np' is not defined

C TypeError: unsupported operand type(s) for +: 'int' and 'str'

D No error, outputs a float value

💡 Hint

Check if input and weight arrays have matching sizes for dot product.

The weights array has length 2 but inputs length 3, so np.dot(weights, inputs) raises a ValueError due to shape mismatch.