Skip to main content
. 2019 Oct 30;12(3):925–941. doi: 10.1111/tops.12474

Figure 3.

Figure 3

Pareto fronts for the ABB (Marcus et al., 1999) learning problem for either one example word (left) or three example words (right). Rightward on x‐axis corresponds to more parsimonious grammars and upward on y‐axis corresponds to grammars that best fit the data, so the best grammars live in the upper right corners of the graphs. Red shade: ground truth grammar. Pink shade: shares structure with ground truth grammar. White shade: incorrect generalizations. As the number of examples increases, the Pareto fronts develop a sharp kink around the ground truth grammar, which indicates a stronger preference for the correct grammar.