Hi, thank you for this interesting work!
How were the precision-recall curves in Figure 7 calculated? Specifically, were thresholds or confidence scores used to generate the precision-recall pairs, or were other methods employed given that the outputs are binary?
Hi, thank you for this interesting work!
How were the precision-recall curves in Figure 7 calculated? Specifically, were thresholds or confidence scores used to generate the precision-recall pairs, or were other methods employed given that the outputs are binary?