Training data and models in Intro to Computing - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

What is the main purpose of training data in machine learning?

Imagine you want to teach a computer to recognize pictures of cats. What role does the training data play in this process?

A It provides examples for the model to learn patterns from.

B It tests how well the model performs on new data.

C It stores the final predictions made by the model.

D It cleans and prepares the data before training.

💡 Hint

Think about what the model needs to see to learn.

Training data is the set of examples the model uses to learn patterns. It helps the model understand what features relate to the target outcome.

🔍 Analysis

intermediate

What is the output of this Python code simulating model training?

Consider this simple code that simulates training a model by counting how many times a value appears in training data.

Intro to Computing

training_data = ['cat', 'dog', 'cat', 'bird', 'dog', 'cat']
model = {}
for item in training_data:
    model[item] = model.get(item, 0) + 1
print(model)

A SyntaxError

B {'cat': 3, 'dog': 2, 'bird': 1}

C {'cat': 3, 'dog': 3, 'bird': 1}

D {'cat': 2, 'dog': 2, 'bird': 1}

💡 Hint

Count how many times each animal appears in the list.

The code counts occurrences of each item in the training data. 'cat' appears 3 times, 'dog' 2 times, and 'bird' once.

❓ data_output

advanced

What is the shape of the training data matrix after preprocessing?

You have 1000 images, each resized to 28x28 pixels and converted to grayscale. The images are stored as a NumPy array for training.

What will be the shape of the training data array?

A (1000, 28, 28)

B (28, 28, 1000)

C (1000, 784)

D (784, 1000)

💡 Hint

Think about flattening each 28x28 image into a single row vector.

Each 28x28 image has 784 pixels. Flattening means each image becomes a 1D array of length 784. With 1000 images, the shape is (1000, 784).

❓ visualization

advanced

Which plot best shows model performance over training epochs?

You trained a model for 10 epochs and recorded accuracy after each epoch. Which plot type best visualizes this change over time?

A Line plot showing accuracy on the y-axis and epochs on the x-axis.

B Bar chart showing accuracy for each class.

C Scatter plot of accuracy vs. number of training samples.

D Pie chart showing percentage of correct predictions.

💡 Hint

Think about how to show change over a sequence of steps.

A line plot is ideal to show how accuracy changes over epochs because it connects points in order and shows trends clearly.

🔍 Analysis

expert

What error does this code raise when training a model with mismatched data?

Look at this code snippet where features and labels have different lengths.

Intro to Computing

features = [[1, 2], [3, 4], [5, 6]]
labels = [0, 1]
from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model.fit(features, labels)

A No error, model trains successfully

B TypeError: unsupported operand type(s) for +: 'int' and 'list'

C IndexError: list index out of range

D ValueError: Found input variables with inconsistent numbers of samples

💡 Hint

Check if the number of feature rows matches the number of labels.

Scikit-learn requires the number of samples in features and labels to be the same. Here, features has 3 samples but labels has 2, causing a ValueError.