The AI project life-cycle typically comprises six core stages: defining the problem, collecting data, preparing data, building the model, evaluating and refining it, and finally, deploying it.
Data gathering focuses on acquiring pertinent data for the defined problem.
Evaluation & Refinements encompass assessing the model's performance and making necessary improvements.
Deployment signifies integrating the model into operational systems for practical application.
Data cleaning is not typically enumerated as a distinct stage, as it is integrated within the broader data preparation phase.
In practice, data cleaning involves error removal, management of missing values, and ensuring data integrity, all performed during data preparation.
Consequently, despite its critical importance, data cleaning is not considered a standalone step among the six primary stages.
Thus, the appropriate selection is option (B) Data cleaning.