| Summary: | The abundance of data along with the significant rise of computational power boost the popularity and usage of machine learning in recent decision making and forecasting. Random forest, being one of the current state of the art model, are well known of its high accuracy and efficiency for both classification and regression problems.
In this paper, Meta Random Forest is introduced to the random forest method to make it even more accurate both in classification and regression problems. Shortly, meta random forest is a method where simple random forests are used as the base for the main random forest. To compare the performance of Meta Random Forest, we use accuracy score for classification problems and R2 score for regression problems.
|
|---|