fasiha/ebisu

Issues

• Numerical instability with high α and β

  #68 opened 5 months ago by fasiha
  0

• Testing with ebisu>=3rc, obtaining a result of probabilityRecall = -1.52. Is this value correct?

  #65 opened 5 months ago by zxl777
  2

• Dev diary: split 3-atom model

  #66 opened a year ago by fasiha
  3

• is it possible to `predict recall` in pure sql?

  #67 opened 9 months ago by LazerJesus
  3

• evaluation

  #22 opened 5 years ago by garfieldnate
  4

• Questions about v3

  #63 opened a year ago by LazerJesus
  17

• Dev diary: single-atom Beta power law

  #64 opened a year ago by fasiha
  11

• Request for comment: a new Ebisu v3 API, the Gamma ensemble

  #62 opened 2 years ago by fasiha
  5

• Request for comment: Ebisu v3 API

  #58 opened 2 years ago by fasiha
  16

• Can this be used as a passive review scheduler?

  #56 opened 2 years ago by c1-g
  9

• Algorithm to select next best item for review (Ebisu for dummies?)

  #57 opened 2 years ago by bunyk
  5

• Mean/var -> Beta moment match can happen in the log-domain

  #7 opened 2 years ago by fasiha
  3

• Better understand probability-based schedules to pick initial a=b

  #16 opened 2 years ago by fasiha
  9

• Look at accuracy of Beta-posterior when rebalancing?

  #18 opened 2 years ago by fasiha
  3

• Simpler expression for predictRecall when b is integer

  #17 opened 2 years ago by fasiha
  2

• Alternate ways of updating after a very surprising quiz result

  #30 opened 2 years ago by fasiha
  0

• Ebisu inertia

  #59 opened 2 years ago by Kaelorn
  3

• Can it be used on micropython?

  #55 opened 3 years ago by AiPacino
  1

• Half-life does not meaningfully increase after reps in some conditions

  #35 opened 4 years ago by brownbat
  17

• Ebisu assumes that half-lives do not change after reviews

  #43 opened 4 years ago by cyphar
  5

• A probability distribution on the halflife itself?

  #32 opened 5 years ago by ttencate
  17

• is the java version of ebisu same as the python one ? or is there Rust version one?

  #53 opened 3 years ago by conanchen
  1

• A surprising difference between q0=0 and q0=0.01

  #52 opened 3 years ago by fasiha
  0

• Support for partial reconsolidation

  #51 opened 3 years ago by jasonsparc
  4

• Modelling Card Correlations

  #27 opened 5 years ago by rasenmaeher92
  3

• non-boolean quiz results

  #23 opened 4 years ago by garfieldnate
  24

• Why not always rebalance?

  #31 opened 4 years ago by ttencate
  6

• Questions from Keybase user

  #42 opened 4 years ago by fasiha
  4

• Comparison to SuperMemo

  #45 opened 4 years ago by BradKML
  0

• How to select the initial parameters?

  #40 opened 4 years ago by ernesto-butto
  11

• Approximate closed-form rebalancing for faster updates and more meaningful half-lives

  #41 opened 4 years ago by matomatical
  2

• Anki Implementation of Ebisu algorithm

  #33 opened 5 years ago by TommasoBendinelli
  10

• Noob Question, should tnow in updateRecall be set to 0 if first time showing the question?

  #37 opened 4 years ago by jahamed
  6

• Typo in formula for Posterior(p|k,n)

  #39 opened 4 years ago by MarcelloNastri
  1

• Typo in math in epsilon-traveled posterior

  #29 opened 4 years ago by fasiha
  1

• How to code output

  #38 opened 4 years ago by elbazjosh
  2

• How difficult do you think it would be to create a C# version of Ebisu?

  #28 opened 4 years ago by rs-mobitech
  4

• Go port

  #10 opened 5 years ago by mustafa0x
  1

• Recall difficulty?

  #19 opened 5 years ago by gcoda
  6

• AssertionError

  #26 opened 5 years ago by tezer
  4

• Ensure EbisuHowto.ipynb runs to completion as part of test/build release

  #13 opened 5 years ago by fasiha
  2

• Upgrading from 1.0.0 to 2.0.0

  #24 opened 5 years ago by killia15
  1

• Minor typo in integral

  #20 opened 5 years ago by fasiha
  2

• Create fresh repo/module for stand-alone pure-Python Ebisu, and a bugfix there

  #21 opened 5 years ago by fasiha
  0

• predictRecallVar missing

  #14 opened 5 years ago by MooersLab
  1

• Last cell of notebook does not run

  #12 opened 5 years ago by MooersLab
  2

• New cards

  #11 opened 5 years ago by mustafa0x
  8

• predictRecall returning negative values

  #8 opened 6 years ago by eshapard
  3

• Simplification of update

  #5 opened 6 years ago by rkern
  10

• predictRecallMedian fails

  #6 opened 6 years ago by fasiha
  1