The optimal model is decided when the error produced by the ensemble learning model is as low as possible, and minimality is decided by the lower value of the loss function. The loss function is a measure of how well a model of prediction can predict the expected result. The most common method of finding the minimum function point is gradient descent. In summary, we must first understand what makes errors in the model to really understand what lies behind an ensemble pattern. We will introduce you briefly to these errors and give each ensemble student an insight into these problems. The error of any model can be mathematically divided into three types.