Hi all,
Exercice M3.01 states:
We will use the following parameters search:
- learning_rate for the values 0.01, 0.1, 1 and 10. This parameter controls the ability of a new tree to correct the error of the previous sequence of trees
- max_leaf_nodes for the values 3, 10, 30. This parameter controls the depth of each tree.
But then the solution uses the following values:
learning_rate = [0.05, 0.1, 0.5, 1, 5]
max_leaf_nodes = [3, 10, 30, 100]
Which leads to a different solution than the one found by solving the problem with the correct values.
The methodology is correct, but this makes it harder to check if one’s solution was correct.
Could a moderator please fix this?