Skip to main content
. 2022 Feb 22;41(8):2100294. doi: 10.1002/minf.202100294

Figure 5.

Figure 5

Cumulative accuracy for different ranks. For single‐label model it corresponds to top‐n accuracy, for multi‐label model it corresponds to rank of the lowest scored ground‐truth divided by the number of ground‐truths. Only the first training run is considered in the graph.