Analytics for the Internet of Things (IoT)

Gradient Boosting Machines (GBM) using R

Similar to random forest, GBM is a decision tree-based model that combines predictions from multiple trees to arrive at an aggregated response. However, instead of thousands of independently grown trees that vote for an answer, GBMs are a series of shallow trees linked together in succession.

Gradient Boosting Machines are a popular and powerful ML technique. A variant of GBMs, called XGBoost, has won several recent Kaggle competitions.