The Altair Community is migrating to a new platform to provide a better experience for you. The RapidMiner Community will merge with the Altair Community at the same time. In preparation for the migration, both communities are on read-only mode from July 15th - July 24th, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here.
Options

How do i proceed with adding cross validation to auto model?

YinYin Member Posts: 17 Contributor II
Here's my attempt at this.

I opened an AM process (i selected SVM), then when i was in the design window, i selected create validation set. After that, i clicked on Split validation and right clicked to replace the operator with CV (under validation). Then i clicked on the CV and in the training window i added a decision tree (DT) and connected the tra from one end to tra in DT and mod in DT to mod on the other end. Then i added apply model and connected the mod to mod and tes to unl from one side and added performance and connected lab in apply model to lab in performance and from the other side of performance i connected per to per (see screenshot). Then i went back out and asked the model to help me solve the problem of connection and then it did this connection which doesn't work, how does one proceed from here?

 

Best Answer

  • Options
    BalazsBaranyBalazsBarany Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert Posts: 955 Unicorn
    Solution Accepted
    Hi!

    You're sending the performance results from the cross validation into two different Remember operators. The comments indicate that these should do something different.

    If you're trying to change an exported AutoModel process for cross validation, this is not the way to do it. You'd include the cross validation on a higher level of the process. But it is probably easier to put it into an new process.

    Write down or copy the results of the AutoModel process: The best learning algorithms, their parameters if you did parameter optimization.

    Then, in a new process, read in your data, possibly do some preprocessing (you might want to take the operators from the exported AutoModel process) and then apply the cross validation with the learning algorithm and the parameters from the AutoModel process. This gives you a much simpler, easier to understand and document process that you can more easily change or reuse.

    Regards,

    Balázs
Sign In or Register to comment.