Model Training Error (H2O)-Gradient Boosted Training

chiaUMchiaUM Member Posts: 4 Newbie
edited February 2021 in Help
Hi everyone,

May I know why when i use the same process to run "random forest" and "decision tree", model the program is able to run.

However, when i use "Gradient Boosted Training" it appears this as shown in Figure below





I have recheck the input parameter max_depth, learn_rate, everything are in order
How can I resolve this? My apologies, im newbie in Machine learning & Rapidminer


Answers

  • tkeneztkenez Employee, RapidMiner Certified Expert, Member Posts: 22 RM Product Management
    Hi @chiaUM ,

    Please double check your Parameter settings in your Optimizie Parameters (Grid) operator. My suspicion is that you try to run some optimization rounds where the parameter values are out of range for GBTs (as the error message hints, the maximal_depth parameter needs to be greater than 0, and the learning_rate needs to be between 0 and 1. Maybe the first iteration of the grid wants to start with e.g. 0 for the max_depth?
  • chiaUMchiaUM Member Posts: 4 Newbie
    Alright, thanks for the information!. would give it a try.
Sign In or Register to comment.