October 2017
Beginner to intermediate
572 pages
26h 1m
English
The errorest function calculates the error evolution of AdaBoost.M1, AdaBoost-SAMME, or the bagging classifiers and returns a vector of error evolutions. In this recipe, we use the boosting and bagging models to generate error evolution vectors and graph the error versus number of trees.
The resulting graph reveals the error rate of each iteration. The trend of the error rate can help measure how fast the errors reduce, while the number of iterations increases. In addition to this, the graphs may show whether the model is over-fitted.
Read now
Unlock full access