A team is planning an ML project and is deciding where to start. In BCS terms, why does the work typically begin with data preparation?
- ABecause the BCS syllabus expressly forbids preparing the data at any point
- BBecause models trained on poorly prepared data produce poor results
- CBecause models do not actually require any data to be prepared for them
- DBecause data preparation can be used to replace the training algorithm itself
An ML project typically begins with data preparation because models trained on poorly prepared data produce poor results.
Why each option is right or wrong.
- B (correct). Poorly prepared data yields poor results, so preparation comes first.
- A. Preparation is required, not forbidden.
- C. Models are built from prepared data, so preparation is required.
- D. Preparation readies the data; it does not replace training.
Reference: BCS Foundation in AI Syllabus v1.2 Module 4 sub-LO 4.1.3.