In simulation, the mean-recall approximation matches the clocked-delay Leitner Queue Network for small arrival rates, and its phase-transition threshold appears to be a conservative lower bound.¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q1
Currency · newest study 2016 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2019: tool and model accuracy changes quickly

Subclaims¶

q1 i? Simulations of 500 reviews over 50 items show the mean-recall approximation performs well for small arrival rates. → Reddy 2016 q1 i? The threshold predicted under the approximation appears to be a lower bound for the true phase transition point in review sessions of moderate length. → Reddy 2016 (2)

Evidence¶

Reddy 2016¶

Reddy, S., Labutov, I., Banerjee, S., & Joachims, T. (2016). Unbounded Human Learning: Optimal Scheduling for Spaced Repetition. KDD ’16, San Francisco, CA, USA. https://doi.org/10.1145/2939672.2939850

q1 · i? · theoretical · r3

Simulation comparing the mean-recall approximation with clocked delays (Fig. 6). "the mean-recall approximation performs well for small values of λext".

"Fig. 6 verifies our intuition that the mean-recall approximation performs well for small values of λext."

Reddy 2016 (2)¶

Reddy, S., Labutov, I., Banerjee, S., & Joachims, T. (2016). Unbounded Human Learning: Optimal Scheduling for Spaced Repetition. KDD ’16, San Francisco, CA, USA. https://doi.org/10.1145/2939672.2939850

q1 · i? · causal · r2

Simulation of the clocked-delay network (Fig. 5) against the approximation's threshold, which "appears to be a lower bound (i.e., a conservative estimate)" for moderate-length sessions.

"The dotted line indicates the phase transition threshold obtained under our mean-recall approximation (Eqn. 4), which appears to be a lower bound (i.e., a conservative estimate) for the true phase transition point for review sessions of moderate length."

Discussion¶

Related Claims¶

• Under the mean-recall approximation, the maximum achievable learning rate is convex in the learner's review frequency budget, suggesting increasing returns at lower budgets. — related
• Under the Leitner Queue Network model, raising the rate of new-item introduction beyond a threshold causes a phase transition in learning rate, which a Mechanical Turk vocabulary experiment reproduced. — related
• In a Mechanical Turk flashcard experiment, raising the new-item arrival rate first increases mastered items, but past the optimum fewer items are mastered and more get stuck in deck 1. — related