Question: When using Extra Tree, only a random subset of the features is considered for splitting at each node ( just as in Random Forest )
When using Extra Tree, only a random subset of the features is considered for splitting at each node just as in Random Forest but in addition, instead of looking for the best threshold for each feature it uses a random threshold for each feature. This increases the randomness in the trees thus lowering the variance and combating overfitting.
True
False
Step by Step Solution
There are 3 Steps involved in it
1 Expert Approved Answer
Step: 1 Unlock
Question Has Been Solved by an Expert!
Get step-by-step solutions from verified subject matter experts
Step: 2 Unlock
Step: 3 Unlock
