In Rapidminer, are the decision tree's weights a measure of the "importance" of attributes in the splitting procedure ?
If yes, why is useful to know these weights ? Are there better methods to know the most discriminant features in a data set ?
Topic rapidminer decision-trees feature-selection data-mining machine-learning
Category Data Science
Basically yes, weights are used for attribute or feature importance. Gradient Boosted Tree weights are the best for feature importance in my tests for most of data. But there are other methods like PCA, Weights by Correlation but the performance would be worse. Also deep learning first two layers (input and first hidden layer) can be used to measure importance of features.
Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.