Improving Molecular Design by Stochastic Iterative Target Augmentation

@inproceedings{Yang2020ImprovingMD,
  title={Improving Molecular Design by Stochastic Iterative Target Augmentation},
  author={Kevin F. Yang and Wengong Jin and Kyle Swanson and R. Barzilay and T. Jaakkola},
  booktitle={ICML},
  year={2020}
}
Generative models in molecular design tend to be richly parameterized, data-hungry neural models, as they must create complex structured objects as outputs. Estimating such models from data may be challenging due to the lack of sufficient training data. In this paper, we propose a surprisingly effective self-training approach for iteratively creating additional molecular targets. We first pre-train the generative model together with a simple property predictor. The property predictor is then… Expand
Break-It-Fix-It: Unsupervised Learning for Program Repair
Learning Space Partitions for Path Planning
  • Kevin Yang, Tianjun Zhang, +6 authors Yuandong Tian
  • Computer Science
  • 2021

References

SHOWING 1-10 OF 51 REFERENCES
Molecular de-novo design through deep reinforcement learning
Execution-Guided Neural Program Synthesis
Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis
Quantifying the chemical beauty of drugs.
Hierarchical Graph-to-Graph Translation for Molecules
Multi-resolution Autoregressive Graph-to-Graph Translation for Molecules
Strategies for Pre-training Graph Neural Networks
...
1
2
3
4
5
...