Options

How to know which attribute (or which is the most important) has the most weight in a neural network

alois_borgognonalois_borgognon Member Posts: 15 Contributor II
edited January 2020 in Help
Hello,

I need to know which attributes are the most important (the most influence/impact on the model) for a price prediction in a neural network model.

I wondered how we could find this.

Thanks a lot for your help !

Aloïs

Answers

  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,517 RM Data Scientist
    Hi,
    Did you check Explain Prediction operator?

    Best,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    alois_borgognonalois_borgognon Member Posts: 15 Contributor II
    no, but I have this kind of model


  • Options
    alois_borgognonalois_borgognon Member Posts: 15 Contributor II

  • Options
    alois_borgognonalois_borgognon Member Posts: 15 Contributor II
    You mean Like this ? Just to know, how to be sure it works or in other words how to set-up explain predictions ?
  • Options
    IngoRMIngoRM Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, Community Manager, RMResearcher, Member, University Professor Posts: 1,751 RM Founder
    Almost, you do not need to apply the model on an extra data set.  In fact, you should provide the original training data of the model to the "tra" port, the model to the "mod" port and a labeled test data set to the "tes" port.
    Hope this helps,
    Ingo
Sign In or Register to comment.