Unable to match split value 10 for feature [V2, 1, 1][2] about numl

[sethjuarez]

Thanks for the code! The problem here is that there is a mathematical combination of features. The decision tree algorithm looks to discriminate based upon single features and branches from there. Essentially since each feature is a random number there is little information gain per feature and therefore does not learn much. I hope this helps!

[sam12github]

what I am aiming to make it learn it is that when it see low degrees it will assume it is weak when it is high degrees it will assume it is high .
that was my main goal from using numl .
thank you for the help .