What is Needed to Make this Random Forest Model Predict into the Future?

SkyTrader
SkyTrader New Altair Community Member
edited November 5 in Community Q&A
Hi, 

I've used the Apply Model operator as suggested to try and make this Random Forest model to predict the "future" -- but it still doesn't predict beyond the last data of my data set (15th Sept 2020)?

I've attached the process to see if anyone can figure out what is needed to get it to make forward future predictions?





Any suggestion much appreciated,

Many thanks.

Answers

  • MartinLiebig
    MartinLiebig
    Altair Employee
    Hi,
    the easiest way would be the lag operator and just shift the label attribute.
    Best,
    Martin
  • SkyTrader
    SkyTrader New Altair Community Member
    Thanks Martin, @mschmitz
    I've had a lot of problems despite re-importing my Excel data using the wizard many times. The attribute is not found anymore (I also had the issue of my Close attribute not being found because of an errant * in the Excel file). Not sure how to proceed as this import / attribute issue has been ongoing for a month now - it used to work when I first started using RM.

    So the end result is I get this:







    I'm assuming I'm using the lag operator correctly? I also tried ticking "include special attributes," 

    Cheers,
  • hughesfleming68
    hughesfleming68 New Altair Community Member
    Hi SkyTrader, I would use also experiment with the windowing operator and sliding window validation. The results of your testing can change dramatically depending on your split point. 

    regards,

    Alex