Fig. 2.

(A–C) Average empirical forgetting rate for the top 25% of pairs in terms of likelihood for MEMORIZE, the uniform reviewing schedule, and the threshold-based reviewing schedule for sequences with different numbers of reviews $n$ and different training periods $T=t_{n-1}-t_1$. Boxes indicate 25% and 75% quantiles and solid lines indicate median values, where lower values indicate better performance. MEMORIZE offers a competitive advantage with respect to the uniform and the threshold-based baselines and, as the training period increases, the number of reviews under which MEMORIZE achieves the greatest competitive advantage increases. For each distinct number of reviews and training periods, $*$ indicates a statistically significant difference (Mann–Whitney U test; P-value < $0.05$) between MEMORIZE vs. threshold and MEMORIZE vs. uniform scheduling.