You are viewing a plain text version of this content. The canonical link for it is here.

Posted to user@ignite.apache.org by dmitrievanthony <dm...@gmail.com> on 2018/06/07 09:58:45 UTC

Re: Xgboost and LSTM support in Apache Ignite ML and DL

Hi, 

    As I've answered on  StackOverflow
<https://stackoverflow.com/questions/50329442/xgboost-and-lstm-support-in-apache-ignite-ml-and-dl/50330775#50330775> 
, 

    Depends on your data location and your goals. Does your data is already
stored in Apache Ignite? Do you use Apache Ignite in your ecosystem? 

    Regarding XGBoost, Apache Ignite doesn't provide trainers for boosted
trees so far, but you can use imported XGBoost models. You can find an
example of such importer here. If you trains your model outside of the
Ignite and then use the model in Ignite to process data this approach will
be good enough for you. 

    Regarding LSTM, Apache Ignite ML currently supports only MLP neural
networks, so you can't train your model inside Apache Ignite. Even so you
still can train the model outside of the Apache Ignite and then import
prepared model. 

Best regards, 
Anton Dmitriev



--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/