In which stage of the CRISP-DM process model is data quality verified for a data mining project?

The stage of the CRISP-DM process model where data quality is verified is during the Data Understanding phase. In this phase, analysts focus on collecting initial data and exploring it to understand its structure, contents, quality, and any underlying patterns. This involves performing data quality assessments to identify issues such as missing values, inaccuracies, or inconsistencies in the data. By addressing these quality concerns early on, data scientists can ensure that the data used for subsequent modeling stages is reliable and relevant, leading to more accurate and effective predictive models.

In contrast, other phases like Data Preparation, Business Understanding, and Evaluation have different focuses. Data Preparation is primarily about transforming and cleaning the data after its quality has already been assessed. Business Understanding involves defining project objectives and requirements, which comes before any data quality checks take place. Evaluation occurs at the end of the process to determine the effectiveness of the model and results, not to verify the data quality.