Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Feature Request]: Automatic Preset Assigning #709

Open
DerIshmaelite opened this issue Nov 27, 2024 · 1 comment
Open

[Feature Request]: Automatic Preset Assigning #709

DerIshmaelite opened this issue Nov 27, 2024 · 1 comment
Labels
enhancement New feature or request

Comments

@DerIshmaelite
Copy link

DerIshmaelite commented Nov 27, 2024

The number of cards in a deck sharing a single preset has a massive influence on RMSE based on personal experience.

  • Having too many cards with varying intrinsic complexities (and difficulty) could decrease the accuracy.
  • Having too few cards on the other hand worsens the RMSE since FSRS needs data to work with.

There must be some theoretical "midpoint" where the number of cards per preset is optimized and the best possible RMSE is achieved.

###ALTERNATIVE

Moreover, I suggest removing "Presets" in general as the method by which FSRS groups cards in order for it to get data and build parameters.

Instead, FSRS should have its card-grouping method, kind of like hidden automated presets which are not shown to the user.
Card-grouping methods could be based on

  1. Lapse-Review Ratio
  2. Accumulated Reviewing Time etc.

I suggest testing this.

@DerIshmaelite DerIshmaelite added the enhancement New feature or request label Nov 27, 2024
@aleksejrs
Copy link

aleksejrs commented Nov 27, 2024

I found that at least some presets with similar minimum recommended retention could be merged without harming RMSE much, but not those with very different MRR.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

No branches or pull requests

2 participants