Skip to content

Selection Algorithms

Octav Popescu edited this page Nov 10, 2026 · 11 revisions

Basically for both mastery algorithms the system computes a score for each candidate problem, as a weighted average of the missing skill probabilities. Where a missing probability is the difference between the mastery level (.95) and the current skill p-known, and the weights are given by the number of opportunities for each skill. And then the Mastery Learning algorithm takes the problem with the highest score, while the Mastery Learning - Easier First one takes the problem with the lowest score. If there are more problems with the same highest/lowest score, they will choose one randomly.

Clone this wiki locally