融合四种视角表征，提升聚合物性质预测精度。
Multi-View Polymer Representations for the Open Polymer Prediction
- 用四种互补表征联合建模：分子指纹、图神经网络、3D信息、预训练SMILES语言模型
- 在NeurIPS 2025挑战赛中公开集MAE达0.057，私有集MAE为0.082
- 适用于需要高精度聚合物性质预测的研究者和工业应用
我们提出一种多视图设计来解决聚合物性质预测问题，充分利用不同表征的互补性。系统整合了四类表征：(i) 表格式的RDKit/Morgan描述符，(ii) 图神经网络，(iii) 3D信息引导的表示，(iv) 预训练的SMILES语言模型，并通过统一集成对每种属性进行平均预测。模型采用10折交叉验证训练，测试时使用SMILES增强。该方法在NeurIPS 2025开放聚合物预测挑战赛中位列2241支队伍中的第9名。提交的集成模型在公共测试集上达到MAE 0.057，私有测试集上为MAE 0.082。
原文摘要 · Abstract (English)
We address polymer property prediction with a multi-view design that exploits complementary representations. Our system integrates four families: (i) tabular RDKit/Morgan descriptors, (ii) graph neural networks, (iii) 3D-informed representations, and (iv) pretrained SMILES language models, and averages per-property predictions via a uniform ensemble. Models are trained with 10-fold splits and evaluated with SMILES test-time augmentation. The approach ranks 9th of 2241 teams in the Open Polymer Prediction Challenge at NeurIPS 2025. The submitted ensemble achieves a public MAE of 0.057 and a private MAE of 0.082.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。