Skip to main content


Fig. 1 | Journal of Cheminformatics

Fig. 1

From: Effect of missing data on multitask prediction methods

Fig. 1

Results of training data sparseness on PKIS data set. a Median RMSD values for DNN (blue) and Macau (red). The light colored traces correspond to the ten sets of hyperparameters, while the dark colored trace is the average of the ten light colored ones. b The RMSD values are scaled relative to the performance of the model on the complete training set; the color scheme is the same as (a)

Back to article page