Recognise the stages of a data analysis project and explain each one

A structured-data machine learning project runs in five steps: conceptualise the modelling task (what the output is and how it will be used), collect the data, prepare and wrangle them, explore them, and train the model. A text-based project follows a parallel path: text problem formulation, data or text curation (for example web scraping or a licensed feed), text preparation and wrangling, text exploration, then model training. The steps are iterative, and results from later steps often send the team back to earlier ones. Questions usually describe an activity and ask which step it belongs to, so tie each activity to its purpose rather than to the tools used.