Training data preparation in Prompt Engineering / GenAI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Why is data normalization important before training a model?

Imagine you have a dataset with features measured in very different units, like height in centimeters and income in dollars. Why should you normalize this data before training a machine learning model?

A Normalization increases the size of the dataset to improve model accuracy.

B Normalization removes missing values automatically from the dataset.

C Normalization ensures all features contribute equally by scaling them to a similar range, preventing bias towards features with larger values.

D Normalization converts categorical data into numerical labels.

💡 Hint

Think about how different scales can affect the learning process of a model.

Normalization scales features to a similar range so that no single feature dominates the learning process due to its scale. This helps models learn better and faster.

❓ Predict Output

intermediate

Output of data splitting code

What is the output of the following Python code that splits data into training and testing sets?

Prompt Engineering / GenAI

from sklearn.model_selection import train_test_split
X = list(range(10))
y = list(range(10))
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
print(len(X_train), len(X_test))

A 3 7

B 7 3

C 10 0

D 0 10

💡 Hint

Check how the test_size parameter affects the split.

The test_size=0.3 means 30% of data goes to testing, so 7 items remain for training and 3 for testing.

❓ Model Choice

advanced

Best data preparation for text classification

You want to train a model to classify movie reviews as positive or negative. Which data preparation step is most important before training?

A Convert text to lowercase, remove punctuation, and tokenize words.

B Normalize numerical features to zero mean and unit variance.

C Fill missing values with the mean of the column.

D Encode categorical variables using one-hot encoding.

💡 Hint

Think about what you do to raw text before feeding it to a model.

Text data needs cleaning like lowercasing, removing punctuation, and tokenizing to convert it into a form the model can understand.

❓ Metrics

advanced

Evaluating data quality impact on model accuracy

You train two models on the same task. Model A uses raw data with many missing values. Model B uses data where missing values were properly handled. Which metric difference would best show the impact of data preparation?

A Model A has lower accuracy and lower loss than Model B.

B Model A has higher accuracy but higher loss than Model B.

C Both models have the same accuracy but different loss values.

D Model B has higher accuracy and lower loss than Model A.

💡 Hint

Good data preparation usually improves both accuracy and loss.

Handling missing values properly improves data quality, leading to better model accuracy and lower loss.

🔧 Debug

expert

Identify the error in data preprocessing code

What error will this Python code raise when preparing data for training?

Prompt Engineering / GenAI

import numpy as np
from sklearn.preprocessing import StandardScaler
X = np.array([[1, 2], [3, 4], [5, 6]])
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)
print(X_scaled[3])

A IndexError: index 3 is out of bounds for axis 0 with size 3

B TypeError: 'StandardScaler' object is not callable

C ValueError: could not convert string to float

D No error, prints scaled values

💡 Hint

Check the size of the array and the index accessed.

The array has only 3 rows (indices 0,1,2). Accessing index 3 causes an IndexError.