OOB Score higher than Validation Score

(Joao Paulo Ramos) #1

Hello guys,

I’m getting the Intro do Machine Learning course, and along with it, trying to apply it’s content to my job.

Well, with that, I’m trying to run a Random Forest Regressor to predict how many days from a certain date a sales opportunity will take to close.

My RFR is getting the following results:

[19.19978950416304, 18.974135023913018, 0.9634958193042105, 0.49285398801014835, 0.9376698492856959]

Where the last 2 values 0.4928 and 0.9376 are the validation score and the OOB score.

My queries are: Should I trust in what score? Both scores should be close to the model “works”?

Many thanks!


(Joao Paulo Ramos) #2

Any update on that? I didn’t found the solution to it.