Which phase transforms raw data into model-ready datasets?

Which phase transforms raw data into model-ready datasets?

• A. Data Understanding
• B. Data Cleaning
• C. Data Preparation
• D. Evaluation

Answer: (C)

Turning raw data into something a model can learn from happens during Data Preparation. This phase involves shaping and transforming the data so it’s clean, consistent, and structured for modeling. It includes cleaning messy data, handling missing values, encoding categorical features, scaling and normalizing numbers, engineering new features, integrating data from multiple sources, and creating training and testing splits. These steps convert messy, unusable data into a dataset that the modeling process can work with effectively. Data Understanding is about exploring the data to learn its properties and potential issues, not about transforming it for modeling. Data Cleaning focuses on fixing or removing bad data, which is part of preparation but doesn’t cover the full set of transformations and feature engineering needed to make the data model-ready. Evaluation is about assessing how well a model performs, not about preparing the data.

Turning raw data into something a model can learn from happens during Data Preparation. This phase involves shaping and transforming the data so it’s clean, consistent, and structured for modeling. It includes cleaning messy data, handling missing values, encoding categorical features, scaling and normalizing numbers, engineering new features, integrating data from multiple sources, and creating training and testing splits. These steps convert messy, unusable data into a dataset that the modeling process can work with effectively.

Data Understanding is about exploring the data to learn its properties and potential issues, not about transforming it for modeling. Data Cleaning focuses on fixing or removing bad data, which is part of preparation but doesn’t cover the full set of transformations and feature engineering needed to make the data model-ready. Evaluation is about assessing how well a model performs, not about preparing the data.