Top left: Confusion matrix for the one-stage classification in the binary classification case (i.e., mergers versus non-mergers), predicted on the dataset in the redshift range 0.52 ≤ z ≤ 0.76. Top right: Similar to the top left figure but for 0.76 ≤ z ≤ 1. Bottom left: Confusion matrix for the two-stage classification in the binary classification case, in the redshift range 0.52 ≤ z ≤ 0.76. Bottom right: Similar to the bottom left figure but for 0.76 ≤ z ≤ 1.