From the course: Applied Machine Learning: Foundations

Unlock the full course today

Join today to access over 25,000 courses taught by industry experts.

Tuning CatBoost

Tuning CatBoost

In this video, we're going to tune our CatBoost model. CatBoost has a bunch of parameters. I'm not going to go over them, but at a high level, these parameters are impacting different aspects of the model. Some of these parameters impact what's called boosting, and you can think of boosting as each subsequent tree. So you can control how many trees there are, you can control the learning rate of the trees. You can think of, if we're going with that golf metaphor, you can think of that learning rate as how hard you're going to hit the ball. Are you going to hit it as hard as you can, or are you going to not hit it quite as hard? And oftentimes, if you hit it as hard as you can, you might hit it off to the side or you might go over. If you just focus on hitting it nice and easy, you might actually do better. And models when you lower the learning rate, often will do better. We also have parameters that deal with the tree. So how deep the tree is, how we grow the tree, similar to the…

Contents