Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.

Recurrent neural networks

prologprolog Member Posts: 4 Contributor I
edited November 2018 in Help
Are recurrent neural networks available in RapidMiner 5?

Answers

  • haddockhaddock Member Posts: 849 Maven
    No.
  • wesselwessel Member Posts: 537 Maven
    What do you want to do with these?
  • prologprolog Member Posts: 4 Contributor I
    wessel wrote:

    What do you want to do with these?
    Well time series prediction.
    Feed-forward NNs can't handle time series prediction well.
  • wesselwessel Member Posts: 537 Maven
    As far as I know this is still active research.
    You have research showing recurrent networks being superior to feed-forward networks for time series forecasting?

    Recurrent neural networks are superior for modelling cognitive processes.
    Since recurrence is proven to be an important part of the workings of our brain.

    If you plan to do research comparing feed-forward networks and recurrent networks, I would strongly recommend to create an own implementation.
  • prologprolog Member Posts: 4 Contributor I
    Recurrent NNs are quite mature already.
    They are several decades old.
    They are like a standard tool to model dynamical systems.

    Feedforward NNs can't extrapolate time series data well. They can only interpolate/do curve-fitting.
    You need recurrent NNs to predict time series generated by dyanamical systems
  • wesselwessel Member Posts: 537 Maven
    prolog wrote:

    Recurrent NNs are quite mature already.
    They are several decades old.
    They are like a standard tool to model dynamical systems.
    For modelling cognitive processes yes.
    prolog wrote:

    Feedforward NNs can't extrapolate time series data well. They can only interpolate/do curve-fitting.
    You need recurrent NNs to predict time series generated by dyanamical systems
    Show me the proof? :P
  • QorrectQorrect Member Posts: 2 Contributor I

    This conversation is already five years old, and there's still no RNN's in Rapid Miner ?  I googled it and this was the first thing to come up ...

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    RNN is not stock in RapidMiner at this very moment. That said we have the DeepLearning4j extension available from our partner, you can check it out here: https://www.rapidminerchina.com/en/products/shop/product/deeplearning4j/

     

  • listslists Member Posts: 39 Maven

    Are there any plans to implement it native- incl. LSTMs?

    Love you all.

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    Not to this moment but since Tensor Flow has an Java API now, maybe some intrepid Java developer could take a stab it and build a Tensor Flow extension. 

  • alihasanalihasan Member Posts: 1 Learner I
    isnot recurrent neural network implemented in rapidminer still??
Sign In or Register to comment.