Benchmark datasets in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding the purpose of benchmark datasets

Why do machine learning researchers use benchmark datasets?

A To train models on random data without labels

B To compare different models fairly using the same data

C To avoid testing models on real-world data

D To increase the size of training data by duplicating samples

💡 Hint

Think about why having a common dataset helps researchers.

Benchmark datasets provide a standard set of data so researchers can compare model performance fairly and consistently.

❓ Predict Output

intermediate

Output of loading a benchmark dataset

What is the output shape of the features when loading the Iris dataset using scikit-learn?

Prompt Engineering / GenAI

from sklearn.datasets import load_iris
iris = load_iris()
X = iris.data
print(X.shape)

A (150, 4)

B (3, 150)

C (150, 3)

D (4, 150)

💡 Hint

Check how many samples and features the Iris dataset has.

The Iris dataset has 150 samples and 4 features, so the shape of the data is (150, 4).

❓ Model Choice

advanced

Choosing a model for the MNIST dataset

You want to classify handwritten digits from the MNIST dataset. Which model is best suited for this task?

A Convolutional Neural Network (CNN)

B Linear Regression

C K-Means Clustering

D Decision Tree Regressor

💡 Hint

Consider the type of data and the task (image classification).

CNNs are designed to work well with image data and are commonly used for digit classification tasks like MNIST.

❓ Metrics

advanced

Evaluating model performance on benchmark datasets

Which metric is most appropriate to evaluate a classification model on the CIFAR-10 benchmark dataset?

A R-squared

B Mean Squared Error

C Accuracy

D Perplexity

💡 Hint

Think about the type of task CIFAR-10 represents.

CIFAR-10 is a classification dataset, so accuracy is a suitable metric to measure how many images are correctly classified.

🔧 Debug

expert

Identifying the error when loading a benchmark dataset

What error will this code raise when trying to load the Boston Housing dataset using scikit-learn?

Prompt Engineering / GenAI

from sklearn.datasets import load_boston
boston = load_boston()

A No error, dataset loads successfully

B ValueError: Dataset not found

C DeprecationWarning: load_boston is deprecated

D ImportError: cannot import name 'load_boston'

💡 Hint

Check recent changes in scikit-learn about this dataset.

The Boston Housing dataset was removed from scikit-learn due to ethical concerns, so trying to import load_boston raises ImportError.