フリー問題

問題文

A retailer wants to split its customers into groups so that each group can receive a different offer. The customer table has purchase history but carries no group labels and nobody has decided how many groups there should be. Which task does this need, and which family of machine learning does that task sit in?

選択肢

1. The task is association, a form of unsupervised learning, so the model finds the links that the carts share instead of a target for each record.
2. The task is regression, a form of supervised learning, so the model finds the line that the totals share instead of a group for each record.
3. The task is classification, a form of supervised learning, so the model finds the edge that the classes share instead of a grouping for each record.
4. The task is clustering, a form of unsupervised learning, so the model finds the qualities that the customers share instead of a target for each record.