From the course: Applied Machine Learning: Foundations
Join today to access over 25,000 courses taught by industry experts.
CatBoost
From the course: Applied Machine Learning: Foundations
CatBoost
In this video, we're going to see how we can improve on our decision tree. Our dummy model is very simple. Linear regression is a little bit more complicated. Decision tree is more complicated than linear regression. One of the cool things about a decision tree is that it can look at a column and make a split based on that column, and then it can come back and look at the column again, which can help it model non-linear relationships between your data and what you're trying to predict. However, we also saw that a decision tree can overfit and in fact will overfit if you use the out-of-the-box settings for it. One of the ways we can get around that is using a model like CatBoost. CatBoost implements what is called a boosted algorithm, and I like to compare this to golf. If we say a decision tree is where I can hit the ball and it's going, the ball will be some distance after it lands away from the hole we're trying to hit it into. With boosting, we can hit the ball once and then we…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.