monaco VS ebisu

At the heart of all serious forecasting is a statistical tool known as Monte Carlo analysis. It allows you to quantify uncertainty by introducing randomness to the inputs of computational models and looking at the range of results. If you want a good example, you might recognize Monte Carlo techniques from Nate Silver’s election forecasts at 538. It's been a gold-standard throughout my career in the space industry, and I can attest to how powerful it is - I've used it to successfully send a rocket to Mars. However, there aren't any tools out there that make it easy for researchers to take their existing models and wrap a Monte Carlo around it. So, I wrote one. It's an open-source python library which I'm calling "monaco". I'm at a point in development where the basic feature set is complete and working well, and I'm looking to finish up the extended roadmap in the next few months. See the project github page for the code, examples, and a lot more info: https://github.com/scottshambaugh/monaco. I’m looking for $1000 to help me present version 1.0 of this tool to the scientific community at the 2022 SciPy Conference in Austin, TX this summer. That amount should cover conference fees, hotel, and airfare, and if you're feeling generous I could use additional funds for some external monitors and cloud compute time. My name is Scott Shambaugh, and if you’re interested in helping fund this please email me at wsshambaugh AT gmail.com. Thank you!

I really wish something like https://github.com/fasiha/ebisu becomes the norm. That is, the idea of fitting the cards to your time (by prioritising) rather than you having to do everything there software wants.

The only bit missing is some algorithm deciding how often to introduce new cards based on your historical data.

It seems from the description that FSRS still puts an exact review date on each card? This feature was pretty much the reason why I stopped using Anki. I'm not in college and not doing exams, I just want to practice when I feel like it, maybe with large breaks between sessions, and not feel like there's a backlog building up.

I think Anki is a great app, I just wish there was an algorithm that would just randomly sample cards (with probability proportional to how urgently you need to review it) rather than put a review date on them. Something like https://github.com/fasiha/ebisu but available as an Anki plugin (if that supports custom algorithms on mobile yet?) or a similar app with an open format for cards.

"GoodNotes uses the Ebisu algorithm for its spaced repetition feature. Ebisu uses a Bayesian model to estimate the probability of remembering a given flashcard, which allows faster adaptation to changes in recall ability. Both algorithms have been shown to be effective in practice, you can learn more about Ebisu at https://fasiha.github.io/ebisu/ "

This is a fundamental issue with SM-2 and how ease factors work. I personally have my Anki settings set up such that there is no ease factor penalty, though I will be working on porting Ebisu v3 to Anki's v3 scheduler once it's ready, which should finally allow us to have proper adaptive ease factors for cards (on all platforms) without the ease hell problem.