Gradient Boosting (CBT)

Build trees one at a time, where each new tree helps to correct errors made by previously trained tree.

Pros:

• Can solve almost anything Cons:
• more sensitive to Overfitting if data is noisy
• training takes longer as trees are built sequentially
• GBMs are harder to tune than RF

https://medium.com/@aravanshad/gradient-boosting-versus-random-forest-cfa3fa8f0d80