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.

[Time Series Forecast] How to put constraints on forecasted value

UseUse Member Posts: 1 Learner I
edited November 2021 in Help

(attached example image for your understanding)


Hello,

I want to know how to put constraints on time series forecast.
For example, I have to forecast 5 kinds of jobs for next 3 years, so I will have 15 forecasted value.
However, there are two constraints on my forecast.

1. Forecasted values have to be more than 0 and less than 100.
2. The sum of each year's forecasted values(the sum of 5 forecasted values of each year) has to be 100, because the values are percentages(ratios).

Is it possible to put these constraints on the forecasted value using RapidMiner?
(I used Holt-Winters.)
My aim is to see how these values work together dynamically, not respectively.

Thank you all and sorry for my bad English  :'(

Best Answers

  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,529 RM Data Scientist
    Solution Accepted
    Hi,
    I suspect the idea of if(value>0,0,value) is not the thing you search for, right?

    ~Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • BalazsBaranyBalazsBarany Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert Posts: 955 Unicorn
    Solution Accepted
    Hi!

    As you have percentages, you usually also should have the base numbers. If you forecast those (and possibly use Martin's method for ensuring that the results are positive), you can automatically calculate the percentages. 

    It depends on your use case, but the expectation is that the forecast should be better for the original values than for the percentages, as the percentages depend more on each other, which the time series forecasting method simply can't know.

    Regards,
    Balázs
Sign In or Register to comment.