Machine Learning Guide for Oil and Gas Using Python

Extra trees (extremely randomized trees)

Before proceeding to gradient boosting, extra trees is covered in this section. Extra trees is another form of supervised ML algorithm that is similar to random forest used for both classification and regression problems. As discussed in the previous section, random forest uses bootstrapping which means samples are drawn with replacement. However, extra trees algorithm uses the whole original sample. In the scikit-learn library, the default term for “bootstrap” is “False” but it does allow for bootstrapping if desired. Another difference between random forest and extra trees is that random forest chooses the optimum split when splitting nodes while extra trees chooses it randomly. Please note that after ...