I have created a random forest algorithm and I am proud of the accuracy obtained.
But I do not understand why if I have 214 rows in the training data and I use 43 number of models (tress) all my trees starts with this 1/1.
Does this number mean that one row has been used to generate that tree? If not, what does this number mean?
Hi @Pili and welcome to the forum.
It’s not clear to me what you’re asking about here. Could you provide some screenshots, and if possible a sample workflow and dataset with a little more information to clarify?
I have a total data set with 286 rows and 25 columns.
The training part has 214 rows.
I optimized the random forest and my best result is using 175 number of trees.
And I used the best parameters. When I take the trees for checking them from the Random Forest learner node, ALL the trees start like the picture with 1/1. I thought that the meaning of this number was the number of trees which follow the decision tree.
I show you the Random Forest learner node configuration.
It’s a little hard only going by screenshots, but I think the problem might be that that you are using a Scorer node after an X-Aggregator node. The X-Aggregator node is intended to replace the Scorer node in the context of a cross validation; you don’t need both nodes.
If you continue to run into problems, please upload a minimal example workflow and dataset that addresses the issue so that I can assist you better.
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.