A data science team at an e-commerce company has developed a highly accurate customer churn prediction model using a complex gradient boosting algorithm. During the Evaluation phase, stakeholders confirm the model's predictive power but state that their primary goal has evolved. They now need to understand the specific reasons why customers are churning to inform retention strategies, a task for which the current "black box" model is ill-suited. According to the CRISP-DM methodology, what is the most appropriate immediate next step?
Return to the Business Understanding phase to redefine the project objectives and success criteria to include model interpretability.
Return to the Data Preparation phase to create new features that might provide more explanatory power when used in a new model.
Return to the Modeling phase to retrain the data with an inherently interpretable model, such as a decision tree or logistic regression.
Proceed to the Deployment phase since the model is technically accurate, and initiate a separate project for root-cause analysis.
The correct answer is to return to the Business Understanding phase. The Cross-Industry Standard Protocol for Data Mining (CRISP-DM) is an iterative process model. The Evaluation phase is designed specifically to assess whether the developed model meets the business success criteria defined in the initial Business Understanding phase. In this scenario, the business goals have fundamentally changed from pure prediction to requiring model interpretability for strategic insights. Because the project's primary objective has been redefined, the team must formally return to the Business Understanding phase to update the project goals, redefine the business success criteria to include interpretability, and adjust the project plan accordingly.
Jumping directly to the Modeling or Data Preparation phases is incorrect because these technical steps should be guided by a clearly defined and agreed-upon business objective. Proceeding to the Deployment phase is also incorrect as it would mean delivering a solution that no longer meets the stakeholders' primary needs, which violates a core principle of the Evaluation phase.
Ask Bash
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
What is CRISP-DM, and why is it important in data projects?
Open an interactive chat with Bash
Why is model interpretability important in certain projects?
Open an interactive chat with Bash
What makes decision trees or logistic regression more interpretable compared to gradient boosting?