Python for AI - Quizzes and Solutions

Test your understanding of each lesson with these comprehensive quizzes. Try to answer all questions before checking the solutions!

Lesson 1: Python Basics for AI

Quiz Questions

1. What is the output of print(10 // 3)?

• [ ] A) 3.33
• [ ] B) 3.0
• [ ] C) 3
• [ ] D) 4

2. Which data type is best for storing model accuracy (e.g., 0.89)?

• [ ] A) int
• [ ] B) float
• [ ] C) string
• [ ] D) bool

3. What does this code print?

loss = 0.0045
print(f"Loss: {loss:.2e}")

• [ ] A) Loss: 0.00
• [ ] B) Loss: 4.50e-03
• [ ] C) Loss: 0.0045
• [ ] D) Loss: 4.5e-3

4. Why should you use / instead of // for calculating accuracy?

• [ ] A) / is faster
• [ ] B) // causes errors
• [ ] C) / preserves decimal precision
• [ ] D) // only works with floats

5. What is the correct way to format a float to 2 decimal places in an f-string?

• [ ] A) f"{value:2f}"
• [ ] B) f"{value:.2f}"
• [ ] C) f"{value:.f2}"
• [ ] D) f"{value:2.f}"

6. What is the output of 2 ** 3?

• [ ] A) 5
• [ ] B) 6
• [ ] C) 8
• [ ] D) 9

7. Which variable name is most appropriate for storing the number of training epochs?

• [ ] A) x
• [ ] B) num_epochs
• [ ] C) EPOCHS
• [ ] D) n

Answers

Lesson 2: Lists and Indexing

Quiz Questions

1. What is the output of [1, 2, 3, 4, 5][2]?

• [ ] A) 1
• [ ] B) 2
• [ ] C) 3
• [ ] D) 4

2. How do you get the last element of a list called data?

• [ ] A) data[last]
• [ ] B) data[-1]
• [ ] C) data[end]
• [ ] D) data[length]

3. What does dataset[:80] return if dataset has 100 elements?

• [ ] A) Elements 0 to 79
• [ ] B) Elements 0 to 80
• [ ] C) Elements 1 to 80
• [ ] D) Elements 80 to 100

4. Which list comprehension creates a list of squares from 0 to 4?

• [ ] A) [x * x for x in range(5)]
• [ ] B) [x ** 2 for x in [0, 1, 2, 3, 4]]
• [ ] C) Both A and B
• [ ] D) Neither A nor B

5. What is the output of len([10, 20, 30])?

• [ ] A) 2
• [ ] B) 3
• [ ] C) 60
• [ ] D) 30

6. How do you create batches of size 16 from a list called data?

• [ ] A) [data[i:i+16] for i in range(len(data))]
• [ ] B) [data[i:i+16] for i in range(0, len(data), 16)]
• [ ] C) data.split(16)
• [ ] D) data[::16]

7. What does [1, 2] + [3, 4] produce?

• [ ] A) [4, 6]
• [ ] B) [1, 2, 3, 4]
• [ ] C) [[1, 2], [3, 4]]
• [ ] D) Error

Answers

Lesson 3: Dictionaries and Data Structures

Quiz Questions

1. How do you access the value associated with key 'lr' in dict config?

• [ ] A) config.lr
• [ ] B) config['lr']
• [ ] C) config(lr)
• [ ] D) config->lr

2. What does config.get('lr', 0.001) return if 'lr' is not in the dict?

• [ ] A) Error
• [ ] B) None
• [ ] C) 0.001
• [ ] D) 'lr'

3. Which data structure should you use to store model hyperparameters?

• [ ] A) List
• [ ] B) Dictionary
• [ ] C) Tuple
• [ ] D) String

4. What is the difference between a tuple and a list?

• [ ] A) Tuples are faster
• [ ] B) Tuples are immutable (cannot be changed)
• [ ] C) Tuples can store any type
• [ ] D) No difference

5. How do you add a new key-value pair to a dictionary?

config = {'lr': 0.001}
# Add 'batch_size': 32

• [ ] A) config.add('batch_size', 32)
• [ ] B) config['batch_size'] = 32
• [ ] C) config.insert('batch_size', 32)
• [ ] D) config.append({'batch_size': 32})

6. What does config.keys() return?

• [ ] A) A list of all values
• [ ] B) A list of all keys
• [ ] C) A view of all keys
• [ ] D) The number of keys

7. Which is faster for storing training metrics across epochs?

• [ ] A) Multiple variables
• [ ] B) List of tuples
• [ ] C) Dictionary with epoch as key
• [ ] D) Nested lists

Answers

Lesson 4: Loops and Iteration

Quiz Questions

1. How many times does this loop run?

for i in range(5):
    print(i)

• [ ] A) 4
• [ ] B) 5
• [ ] C) 6
• [ ] D) Infinite

2. What does enumerate([10, 20, 30]) provide?

• [ ] A) Just the values
• [ ] B) Just the indices
• [ ] C) Index-value pairs
• [ ] D) Length of the list

3. How do you iterate through two lists simultaneously?

• [ ] A) for a, b in zip(list1, list2):
• [ ] B) for a in list1 and b in list2:
• [ ] C) for (a, b) in [list1, list2]:
• [ ] D) Cannot be done

4. What does break do in a loop?

• [ ] A) Skips the current iteration
• [ ] B) Exits the loop entirely
• [ ] C) Pauses the loop
• [ ] D) Restarts the loop

5. What is the output?

for i in range(3):
    if i == 1:
        continue
    print(i)

• [ ] A) 0 1 2
• [ ] B) 0 2
• [ ] C) 1
• [ ] D) 0 1

6. How do you create a training loop for 10 epochs?

• [ ] A) for epoch in range(10):
• [ ] B) while epoch < 10:
• [ ] C) Both A and B
• [ ] D) Neither A nor B

7. What is the benefit of using enumerate() over manual indexing?

• [ ] A) Faster execution
• [ ] B) Cleaner, more readable code
• [ ] C) Works with more data types
• [ ] D) No benefit

Answers

for epoch in range(num_epochs):
    for batch_idx, batch_data in enumerate(dataloader):
        # Training step
        loss = train_step(batch_data)
        if loss < threshold:
            break  # Early stopping

Lesson 5: NumPy Essentials

Quiz Questions

1. Why is NumPy faster than Python lists?

• [ ] A) Written in C
• [ ] B) Uses vectorized operations
• [ ] C) Stores data in contiguous memory
• [ ] D) All of the above

2. What is the shape of np.zeros((3, 4))?

• [ ] A) (3,)
• [ ] B) (4,)
• [ ] C) (3, 4)
• [ ] D) (4, 3)

3. What does broadcasting mean in NumPy?

• [ ] A) Copying arrays
• [ ] B) Automatically matching array shapes for operations
• [ ] C) Sending data over network
• [ ] D) Flattening arrays

4. How do you reshape a 1D array of 12 elements into (3, 4)?

• [ ] A) arr.reshape(3, 4)
• [ ] B) arr.shape = (3, 4)
• [ ] C) Both A and B
• [ ] D) Cannot be done

5. What is the output of np.array([1, 2, 3]) * 2?

• [ ] A) [1, 2, 3, 1, 2, 3]
• [ ] B) [2, 4, 6]
• [ ] C) 6
• [ ] D) Error

6. How do you access the first row of a 2D array arr?

• [ ] A) arr[0]
• [ ] B) arr[0, :]
• [ ] C) Both A and B
• [ ] D) arr[:, 0]

7. What shape represents a batch of 32 RGB images of size 64x64?

• [ ] A) (32, 64, 64, 3)
• [ ] B) (64, 64, 3, 32)
• [ ] C) (3, 64, 64, 32)
• [ ] D) (64, 64, 32, 3)

Answers

Lesson 6: Functions and Modules

Quiz Questions

1. How do you define a function that calculates accuracy?

• [ ] A) function accuracy(preds, targets):
• [ ] B) def accuracy(preds, targets):
• [ ] C) func accuracy(preds, targets):
• [ ] D) create accuracy(preds, targets):

2. What does a function without a return statement return?

• [ ] A) 0
• [ ] B) None
• [ ] C) Error
• [ ] D) Empty string

3. How do you import only the nn module from PyTorch?

• [ ] A) import torch.nn
• [ ] B) from torch import nn
• [ ] C) Both A and B
• [ ] D) import nn from torch

4. What is a lambda function?

• [ ] A) A named function
• [ ] B) An anonymous one-line function
• [ ] C) A class method
• [ ] D) A built-in function

5. What does *args allow in a function definition?

• [ ] A) Variable number of positional arguments
• [ ] B) Variable number of keyword arguments
• [ ] C) Required arguments
• [ ] D) Default arguments

6. How do you set a default value for a parameter?

def train(epochs=10):
    pass

• [ ] A) Correct
• [ ] B) def train(epochs: 10):
• [ ] C) def train(epochs -> 10):
• [ ] D) def train(epochs ? 10):

7. What is the purpose of docstrings?

• [ ] A) Execute code
• [ ] B) Document function purpose and parameters
• [ ] C) Import modules
• [ ] D) Define variables

Answers

def calculate_loss(predictions, targets, reduction='mean'):
    """
    Calculate loss between predictions and targets.

    Args:
        predictions: Model outputs
        targets: Ground truth labels
        reduction: 'mean', 'sum', or 'none'

    Returns:
        Loss value as float
    """
    # Implementation here
    pass

Lesson 7: Mini-Project - MNIST

Project Checklist

Use this checklist to verify your MNIST project is complete:

Data Loading: - [ ] Successfully loaded MNIST dataset - [ ] Verified data shape (8x8 images, 1797 samples) - [ ] Checked for missing values

Preprocessing: - [ ] Normalized pixel values to [0, 1] range - [ ] Flattened images from 2D to 1D - [ ] Split into training and test sets (80/20)

Model Training: - [ ] Initialized Logistic Regression classifier - [ ] Trained model on training data - [ ] Achieved >90% test accuracy

Evaluation: - [ ] Calculated accuracy on test set - [ ] Created confusion matrix - [ ] Identified most common misclassifications

Visualization: - [ ] Plotted sample digits from dataset - [ ] Visualized confusion matrix as heatmap - [ ] Showed examples of correct and incorrect predictions

Code Quality: - [ ] Used descriptive variable names - [ ] Added comments explaining steps - [ ] Created reusable functions - [ ] Included error handling

Challenge Questions

1. What accuracy did your model achieve?

Expected: >90% (typically 95-97% for MNIST digits dataset)

2. Which digit pairs are most often confused?

Common: 4 ↔ 9, 3 ↔ 8, 5 ↔ 3, 7 ↔ 1

3. How can you improve model accuracy?

• [ ] A) Use more training data
• [ ] B) Try different classifiers (SVM, Random Forest)
• [ ] C) Feature engineering (gradients, edges)
• [ ] D) All of the above

Answer: D) All of the above

4. What is the computational advantage of Logistic Regression?

• Fast training (seconds vs minutes for deep learning)
• No GPU required
• Simple, interpretable model
• Good baseline for comparison

Final Assessment

Score Your Project:

• Beginner (50-70%): Basic loading and training works
• Intermediate (70-85%): Includes visualization and evaluation
• Advanced (85-95%): Clean code, functions, error handling
• Expert (95-100%): Additional features (cross-validation, hyperparameter tuning)

Overall Course Assessment

Self-Evaluation Checklist

After completing all 7 lessons, evaluate your skills:

Python Fundamentals: - [ ] I can create and manipulate variables of different types - [ ] I understand when to use / vs // - [ ] I can format strings using f-strings - [ ] I write clear, commented code

Data Structures: - [ ] I can work with lists, dictionaries, and tuples - [ ] I understand indexing and slicing - [ ] I can use list comprehensions - [ ] I can store model configurations in dictionaries

Control Flow: - [ ] I can write for and while loops - [ ] I understand enumerate, zip, and range - [ ] I can use break and continue effectively - [ ] I can implement training loops

NumPy: - [ ] I understand the difference between lists and arrays - [ ] I can create and reshape NumPy arrays - [ ] I understand broadcasting - [ ] I can work with multi-dimensional arrays

Functions: - [ ] I can define and call functions - [ ] I can use default arguments - [ ] I understand lambda functions - [ ] I can import and use external libraries

Project Skills: - [ ] I can load and preprocess data - [ ] I can train a machine learning model - [ ] I can evaluate model performance - [ ] I can visualize results

Next Steps Based on Score

Checked 0-20 boxes (0-33%): - Re-watch videos and re-read lessons - Practice more in Colab notebooks - Ask for help in GitHub Discussions

Checked 21-40 boxes (34-66%): - Review weak areas - Complete additional practice exercises - Build a personal project

Checked 41-50 boxes (67-83%): - Good understanding! Move to next course - Start exploring PyTorch/TensorFlow - Contribute to open-source projects

Checked 51-60 boxes (84-100%): - Excellent mastery! You're ready for deep learning - Begin "Introduction to ML" course - Share your project on GitHub

Additional Practice Problems

Challenge 1: Data Preprocessing Pipeline

Create a function that: 1. Takes a list of training samples 2. Normalizes values to [0, 1] 3. Splits into train/val sets 4. Returns both sets as NumPy arrays

Challenge 2: Training Logger

Build a class that: 1. Stores metrics for each epoch 2. Calculates running averages 3. Prints formatted logs 4. Exports results to CSV

Challenge 3: Hyperparameter Grid Search

Implement a script that: 1. Defines a grid of hyperparameters 2. Trains models for each combination 3. Tracks best-performing configuration 4. Visualizes results

Need Help?

• 📖 Python Official Documentation
• 💬 GitHub Discussions
• 🎥 Additional Video Resources
• 📧 Contact: Submit an issue on GitHub

← Back to Course Home | Continue to Math Essentials →