AI RESEARCH
Multi-RF Fusion with Multi-GNN Blending for Molecular Property Prediction
arXiv CS.AI
•
ArXi:2603.20724v1 Announce Type: new Multi-RF Fusion achieves a test ROC-AUC of 0.8476 +/- 0.0002 on ogbg-molhi (10 seeds), placing on the OGB leaderboard ahead of HyperFusion (0.8475 +/- 0.0003). The core of the method is a rank-averaged ensemble of 12 Random Forest models trained on concatenated molecular fingerprints (FCFP, ECFP, MACCS, atom pairs -- 4,263 dimensions total), blended with deep-ensembled GNN predictions at 12% weight.