Fig. 10
From: Human-in-the-loop active learning for goal-oriented molecule generation
Double sigmoid transformation function used in use-case 1 experiment to map predicted LogP values to [0, 1] scores
Fig. 10 | Journal of Cheminformatics