This use-case focuses on building and testing the performance of computational models for predicting power conversion efficiency of OPVs based on the SMILES-derived structural information of the donor candidates as well as assessing impact and implications of the choice of training data: large but synthetic vs small but experimental datasets.

Link.