Exams / AIF-C01 / Use cases and ML techniques / Question 3
AIF-C01 practice question: Use cases and ML techniques (3 of 5)
Fundamentals of AI and ML A free question from our original AIF-C01 practice exam. Try it, then open the answer to see why each option is right or wrong.
A payroll team must calculate employee tax deductions using rates that are fixed by law and published every year. Every result must be exactly reproducible for auditors. A manager suggests training an ML model on last year's payslips. What is the BEST advice?
Choose one answer.
- AA. Train a deep learning model, because it will learn the tax rules more precisely
- BB. Use a foundation model, because it can read the published tax tables directly
- CC. Use rule-based code, because the answer is deterministic and the rules are known
- DD. Train a clustering model, because it will group employees into tax bands
Show the answer and explanation
Answer: C
- A. Wrong: A model learns approximate patterns and can make errors. It cannot guarantee the exact, legally defined answer.
- B. Wrong: A foundation model is nondeterministic and can hallucinate figures, which fails the reproducibility requirement.
- C. Right: Correct. When the logic is fully known and an exact answer is required, ML adds cost and risk without adding value.
- D. Wrong: Tax bands are already defined by law. Discovering groups from data adds nothing and could misplace employees.
Explanation
ML is valuable when patterns are too complex to write as rules or when the rules are unknown. When a problem has fixed, published rules and needs a deterministic answer every time, conventional code is cheaper, simpler and more reliable. Recognising when AI is not the right tool is part of the exam.
Study
Keep practising
Answer questions with a score at the end, or practise just this topic with the explanation after each one. No sign-up needed.