Open pratikgandhic1 opened 1 year ago
This would be an open research question, since the original paper doesn't cover it, and depends on the properties of the dataset you are trying to filter.
The best way to go about this is through experimentation - you should have an idea of what "good" means (either through labelling, or by using another outlier algorithm that acts with "perfect knowledge" of the previous batch data), and compare various values against that benchmark.
Hi there,
I had a general question: What is the usual criteria of selecting the number of trees and tree_size in a RCF model?
Thanks!