Training data preparation in Prompt Engineering / GenAI - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is training data preparation in machine learning?

Training data preparation is the process of cleaning, organizing, and formatting raw data so that a machine learning model can learn from it effectively.

beginner

Why do we need to clean data before training a model?

Cleaning data removes errors, missing values, and inconsistencies that could confuse the model and reduce its accuracy.

intermediate

What is feature scaling and why is it important?

Feature scaling adjusts the range of data features so they have similar scales, helping the model learn faster and perform better.

beginner

Explain the difference between training, validation, and test data.

Training data is used to teach the model. Validation data helps tune the model’s settings. Test data checks how well the model works on new, unseen data.

intermediate

What is data augmentation and when is it used?

Data augmentation creates new training examples by modifying existing data, like flipping images. It is used to increase data size and improve model robustness.

Which step is NOT part of training data preparation?

A Cleaning missing values

B Splitting data into sets

C Training the model

D Scaling features

Training the model happens after data preparation, not during it.

Why do we split data into training, validation, and test sets?

A To evaluate model performance fairly

B To remove errors from data

C To make the dataset smaller

D To speed up data cleaning

Splitting data helps test how well the model works on new data and tune it properly.

What does feature scaling do?

A Adds new data points

B Removes missing data

C Splits data into groups

D Changes data to a similar range

Feature scaling adjusts data values to a similar range for better model learning.

Data augmentation is mainly used to:

A Create more training examples

B Clean data errors

C Split data into sets

D Scale features

Data augmentation increases the size of training data by creating new examples.

Which of these is a common data cleaning task?

A Normalizing features

B Removing duplicates

C Splitting data

D Training the model

Removing duplicate records is a common cleaning step to improve data quality.

Describe the key steps involved in preparing training data for a machine learning model.

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

Explain why splitting data into training, validation, and test sets is important.

Consider how you check if a model works well on new data.