Skip to main content
. 2021 Jul 21;2(8):100312. doi: 10.1016/j.patter.2021.100312

Figure 3.

Figure 3

Private leaderboard log loss and AUC for the top ten ranking teams and the baseline model

The error bars correspond to the 95% confidence intervals. The top two teams outperformed all other submissions in the leaderboard challenge as well as the baseline model.