Pre-requisites

• Regression Trees
• Gradient Boost

The first step in XG Boost is to make an intial prediction. This prediction can be anything. The second step is to calculate the error. This error is the difference between the prediction and the actual value.