August 2019
Intermediate to advanced
342 pages
9h 35m
English
We have seen that decision trees suffer from some important limitations, which can lead to unstable results that are caused even by small variations in the training data. To improve forecasts, you can use ensemble algorithms, such as random forest.
Random forest is nothing but a decision tree ensemble in which each tree is given a vote. The improvement in forecasts is consequently determined by the count of the votes attributed to them: the forecasts that obtain the highest number of votes are those that are selected to achieve the final result of the algorithm.
The creator of the Random Forest algorithm, Leo Breiman, noted that the results obtained by an ensemble of trees improved if the trees ...
Read now
Unlock full access