フリー問題

問題文

A team has seen the similarity score of its model drop, and is also reading how two of the rating factors are defined. The first line of the note has to name the simplest way to correct that drop. The second line has to set out which factor weighs a representative training set and which weighs the spread of predictions. Which pairing belongs in the note?

選択肢

1. The simplest way to correct a drop in the similarity score is rebalance, and the rating weighs balance for a representative set and coverage for the spread of predictions.
2. The simplest way to correct a drop in the similarity score is recasting, and the rating weighs coverage for a representative set and balance for the spread of predictions.
3. The simplest way to correct a drop in the similarity score is recasting, and the rating weighs balance for a representative set and coverage for the spread of predictions.
4. The simplest way to correct a drop in the similarity score is rebalance, and the rating weighs coverage for a representative set and balance for the spread of predictions.