SOA Exam SRM Sample Question 69
Study guide: Exam SRM
You are fitting a boosted regression tree with shrinkage parameter \lambda = 0.08. There is a single numeric predictor variable, x, and the target variable is y. Each tree is to have a single split. The first tree, built using all the training data has the following split:
- If x < 17.5, then the predicted value of y is 125.4.
- If x ≥ 17.5, then the predicted value of y is 350.5.
The following are three observations from the training set:
| Observation | x | y |
|---|---|---|
| 1 | 15.2 | 139 |
| 2 | 12.9 | 156 |
| 3 | 23.6 | 289 |
Calculate the updated predictions for these three observations (\hat{f}(x_1), \hat{f}(x_2), \hat{f}(x_3)).
- A) (–406.1, –389.1, –31.1)
- B) (–211.4, –194.4, 163.6)
- C) (10.0, 10.0, 28.0)
- D) (111.0, 128.0, 279.0)
- E) (129.0, 146.0, 261.0)
Answer this question and see the worked solution