Options

Auto model process

micheljanosmicheljanos Member Posts: 40 Maven
edited June 2019 in Help

1. I'm trying auto model in a large dataset that takes a lot of time to process. I already know some parameters I want in the GBT model, but I cannot (don't know) acess the process because the "open process" doesn't open before the process is finished. 

2. If this GBT is the XGB implementation why it takes so long?

Best,

 

Michel

Tagged:

Answers

  • Options
    sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager

    tagging @IngoRM

  • Options
    yyhuangyyhuang Administrator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 364 RM Data Scientist

    Hi  @micheljanos,

     

    There exist several implementations of the GBDT family of model such as: GBM, XGBoost, LightGBM, Catboost, etc.

    The GBT model in RM studio is integrated from H2O library for GBM, and as you may know training GBM is slow. 

     

    The xgboost is defintely the GBM killer but unfortunately we have not integrated into RapidMiner yet. 

    I am sorry for not giving you a specific and full answer about the mathematical difference between GBM and XGboost, for more details, please have a read of two papers

     

    https://arxiv.org/pdf/1603.02754.pdf

    https://statweb.stanford.edu/~jhf/ftp/trebst.pdf

     

    All the best,

    YY

Sign In or Register to comment.