Multi-class classification in ML Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding the output shape of a multi-class classification model

You build a neural network for a multi-class classification problem with 5 classes. The last layer uses a softmax activation. What should be the shape of the output layer?

A Five neurons with softmax activation

B A single neuron with sigmoid activation

C Five neurons with sigmoid activation

D A single neuron with softmax activation

💡 Hint

Think about how the model outputs probabilities for each class.

For multi-class classification, the output layer should have as many neurons as classes. Softmax activation converts outputs into probabilities that sum to 1.

❓ Metrics

intermediate

Choosing the right metric for multi-class classification

You trained a multi-class classifier with 4 classes. Which metric is best to evaluate overall model performance when classes are imbalanced?

A Precision for class 1 only

B Macro-averaged F1 score

C Accuracy

D Mean Squared Error

💡 Hint

Consider a metric that treats all classes equally regardless of their frequency.

Macro-averaged F1 score calculates F1 for each class and averages them, giving equal importance to all classes, which is good for imbalanced data.

❓ Predict Output

advanced

Output of softmax probabilities for multi-class prediction

What is the output of the following Python code snippet?

ML Python

import numpy as np
from scipy.special import softmax

logits = np.array([2.0, 1.0, 0.1])
probabilities = softmax(logits)
print(probabilities)

A [0.73105858 0.19661193 0.07232949]

B [0.50000000 0.30000000 0.20000000]

C [0.65900114 0.24243297 0.09856589]

D [0.33333333 0.33333333 0.33333333]

💡 Hint

Softmax converts logits to probabilities that sum to 1.

The softmax function exponentiates each logit and normalizes by the sum of all exponentials. The output sums to 1 and reflects relative confidence.

❓ Hyperparameter

advanced

Choosing the right loss function for multi-class classification

You are training a neural network for a 3-class classification problem. Which loss function should you use to train the model effectively?

A Categorical Cross-Entropy

B Binary Cross-Entropy

C Hinge Loss

D Mean Squared Error

💡 Hint

Consider a loss function designed for multi-class problems with one-hot labels.

Categorical Cross-Entropy is the standard loss for multi-class classification with one-hot encoded labels and softmax output.

🔧 Debug

expert

Identifying the cause of poor multi-class classification accuracy

You trained a multi-class classifier with 6 classes. The training accuracy is high, but test accuracy is very low. Which of the following is the most likely cause?

A Incorrect use of sigmoid activation in output layer

B Model is underfitting due to too few epochs

C Using categorical cross-entropy loss with one-hot labels

D Model is overfitting due to lack of regularization

💡 Hint

High training accuracy but low test accuracy usually means the model memorizes training data.

Overfitting happens when the model learns training data too well but fails to generalize. Lack of regularization can cause this.