Tree-Based ML Models

Random Forest, XGBoost, and More

Random Forest

The random forest algorithm has been around since the 90s, and according to Kaggle surveys, …

Beware Default Random Forest Importances

Important reading for anyone using linear model coefficients or tree-based feature importance from Terence Parr et al.

LinkedIn Post - Random Forest Hyperparameters

A note of caution about scikit-learn’s default settings

XGBoost

The XGBoost algorithm is slightly more recent.

XGBoost Resources

Series of four detailed videos about how XGBoost works from Josh Starmer at StatQuest

Other Tree-Based Models

Other models that use an ensemble of decision trees include gradient boosting. This article below uses a gradient boosting machine (GBM) for time series prediction.

How DoorDash Improves Holiday Predictions via Cascade ML Approach

A real example of time series forecasting from Chad Akkoyun and Zainab Danish at DoorDash