SynCoTrain uses co-training of SchNet and ALIGNN under positive-unlabeled learning to assign synthesizability labels to theoretical oxide crystals, reporting 95-97% recall on known positives.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.mtrl-sci 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SynCoTrain: A Dual Classifier PU-learning Framework for Synthesizability Prediction
SynCoTrain uses co-training of SchNet and ALIGNN under positive-unlabeled learning to assign synthesizability labels to theoretical oxide crystals, reporting 95-97% recall on known positives.