Skip to content

How to predict ranking on new data? #19

@supercar1

Description

@supercar1

Hello again! I've posted here before (#18), but my explanation of the issue was not good enough, and as the issue was closed (and not solved), I had no other option than to submit a new issue.

What I'm trying to understand, is how to rank a "new" dataset, that wasn't used for training, validation or testing.
What I mean is that I can run the code, and I believe a ranking model is created (.pkl file), but that is based on the data that was used for training. What if I want to rank new data, that I can't even use for training because I don't know the "right" rank order? Or is this not possible with ptranking?

I hope I could explain myself better now, and thanks in advance!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions