From the course: Applied Machine Learning: Supervised Learning
Join today to access over 24,900 courses taught by industry experts.
Challenge: XGBoost model
From the course: Applied Machine Learning: Supervised Learning
Challenge: XGBoost model
- [Instructor] Okay, it's challenge time again. My challenge for you is to build an XGBoost classifier with the Titanic dataset and then use grid search on it to see if you can beat the out-of-the-box model. Some hyper parameters that you might want to look at adjusting are the number of estimators. Again, this is the number of times you can hit the ball in golfing terms, the max depth and the learning rate. I'm going to give you a hint. Oftentimes I find that XGBoost will overfit out-of-the-box, and so one of the ways to make it not overfit is to lower that depth. I think the max depth default is six, which often will make the model overfit. Good luck.
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.