aiqbench analysis
CalibHyper introduces a novel chance-corrected relational hypergraph approach to address data scarcity in molecular property prediction, a critical bottleneck for drug discovery and materials science. By refining how auxiliary labels are weighted through joint distribution analysis rather than simple agreement metrics, the method achieves competitive performance in few-shot settings without relying on expensive experimental data. This advancement directly supports the development of more efficient machine learning models for scientific discovery where labeled datasets are inherently limited.
CalibHyper: Chance-Corrected Relational Hypergraphs for Few-Shot Molecular Property Prediction
arXiv:2609.33342v1 Announce Type: new Abstract: Molecular property prediction is central to drug development and materials discovery, but experiments are costly and labeled data are scarce. Context-aware methods use auxiliary assay labels to support few-shot prediction, and recent work supervises property relations with label agreement. However, label agreement is sensitive to class marginals and does not directly capture dependence between properties. We propose CalibHyper, a chance-corrected relational hypergraph method based on the joint label distribution. CalibHyper subtracts an independence baseline from the ordered four-state label distribution and shrinks the residual according to the number of joint observations. A swap-equivariant relation head estimates these residuals, which choose the auxiliary properties for each molecule and set the sign and weight of their hyperedge messages. On thirteen datasets from five benchmarks, in both 1-shot and 10-shot settings, CalibHyper and its ablation settings achieve ROC-AUC competitive with the strongest reported results.
Abstract:Molecular property prediction is central to drug development and materials discovery, but experiments are costly and labeled data are scarce. Context-aware methods use auxiliary assay labels to support few-shot prediction, and recent work supervises property relations with label agreement. However, label agreement is sensitive to class marginals and does not directly capture dependence between properties. We propose CalibHyper, a chance-corrected relational hypergraph method based on the joint label distribution. CalibHyper subtracts an independence baseline from the ordered four-state label distribution and shrinks the residual according to the number of joint observations. A swap-equivariant relation head estimates these residuals, which choose the auxiliary properties for each molecule and set the sign and weight of their hyperedge messages. On thirteen datasets from five benchmarks, in both 1-shot and 10-shot settings, CalibHyper and its ablation settings achieve ROC-AUC competitive with the strongest reported results.
Machine Learning (cs.LG); Computation and Language (cs.CL)
arXiv:2609.33342 [cs.LG]
(or arXiv:2609.33342v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.33342 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 239 words
