MLS-C01 Question 132
Select 2A retail company wants to improve its product recommendation system. Currently, they use a simple rule-based approach where customers are shown products frequently purchased together. The business team is considering building a machine learning (ML) model to replace the current system. Under which of the following conditions should the company NOT use ML?
- A
The existing rule-based system is already achieving the desired business outcomes, and there is no significant need for improvement.
- B
The company does not have enough historical transactional data to train a machine learning model effectively.
- C
The company has access to highly skilled data scientists and sufficient computational resources to build and deploy an ML model.
- D
The business requirements demand real-time recommendations for millions of customers, and the team has already implemented an ML-based solution that meets these needs.
- E
The company plans to integrate personalized recommendations based on customer behavior and preferences, which are not easily captured by simple rules.
Show answer and explanation
Correct answers: A, B
Explanation
Machine learning should be used only when it provides a clear advantage over existing solutions. If the current system is already effective or if there is insufficient data to train a model, ML may not be the right choice. The decision to use ML should always be based on evaluating the problem requirements, available data, and expected outcomes.
- A. Correct.
If the current rule-based system is already achieving the desired outcomes, there may not be a strong justification for investing in a machine learning solution. ML should only be used when it provides clear value over existing methods.
- B. Correct.
Machine learning models require sufficient data to train effectively. Without enough historical data, it is unlikely that an ML model would perform well, making it an unsuitable solution in this scenario.
- C. Incorrect.
Having skilled data scientists and computational resources is a prerequisite for implementing ML, but it does not necessarily mean ML should be used. The decision depends on the problem and the business needs.
- D. Incorrect.
If an ML-based solution is already implemented and meeting business requirements, there is no need to reconsider using ML for this use case. This option does not indicate a scenario where ML is inappropriate.
- E. Incorrect.
Personalized recommendations based on customer behavior and preferences often require ML to analyze patterns and make predictions. This scenario is a good use case for ML.