Computer scienceData scienceMachine learningEnsemble learning

Gradient boosting

Theory

The tree depth control

Report a typo

Suppose we trained gradient boosting with the learning rate of 0.1 for 800 rounds in the regression setting. We want to see if setting different maximum tree depths will help to improve the ensemble performance:

A box plot of 1,2,6, and 10 maximum tree depths, with maximum depth on the X axis and MSE on the Y axis

Based on the plot above, what is the optimal value for the maximum tree depth?

Select one option from the list
___

Create a free account to access the full topic