Abstract
Data imbalance is easily found in annotated data when the observations of certain continuous label values are difficult to collect for regression tasks. When they come to molecule and polymer property predictions, the annotated graph datasets are often small because labeling them requires expensive equipment and effort. To address the lack of examples of rare label values in graph regression tasks, we propose a semi-supervised framework to progressively balance training data and reduce model bias via self-training. The training data balance is achieved by (1) pseudo-labeling more graphs for under-represented labels with a novel regression confidence measurement and (2) augmenting graph examples in latent space for remaining rare labels after data balancing with pseudo-labels. The former is to identify quality examples from unlabeled data whose labels are confidently predicted and sample a subset of them with a reverse distribution from the imbalanced annotated data. The latter collaborates with the former to target a perfect balance using a novel label-anchored mixup algorithm. We perform experiments in seven regression tasks on graph datasets. Results demonstrate that the proposed framework significantly reduces the error of predicted graph properties, especially in under-represented label areas. © 2023 ACM.
| Original language | English |
|---|---|
| Title of host publication | KDD '23 - Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining |
| Place of Publication | New York, NY |
| Publisher | Association for Computing Machinery |
| Pages | 1453-1465 |
| Number of pages | 13 |
| ISBN (Print) | 9798400701030 |
| DOIs | |
| Publication status | Published - Aug 2023 |
| Externally published | Yes |
| Event | 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2023) - Long Beach Convention & Entertainment Center, Long Beach, United States Duration: 6 Aug 2023 → 10 Aug 2023 https://kdd.org/kdd2023/ |
Publication series
| Name | Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining |
|---|---|
| ISSN (Print) | 2154-817X |
Conference
| Conference | 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2023) |
|---|---|
| Abbreviated title | KDD ’23 |
| Place | United States |
| City | Long Beach |
| Period | 6/08/23 → 10/08/23 |
| Internet address |
Funding
This work was supported in part by NSF IIS-2142827, IIS-2146761, CBET-2102592, and ONR N00014-22-1-2507.
Research Keywords
- graph property regression
- imbalanced regression
- self-training
- semi-supervised learning
Fingerprint
Dive into the research topics of 'Semi-Supervised Graph Imbalanced Regression'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver