Weighted Ensemble Approach for Knowledge Graph completion improves performance
Sinha, M.; Tu, R.; Gonzalez, C.; Su, A. I.
Show abstract
This study introduces a weighted ensemble method called "WeightedKgBlend" for link prediction in knowledge graphs which combines the predictive capabilities of two types of Knowledge Graph completion methods: knowledge graph embedding and path based reasoning. By dynamically assigning weights based on individual model performance, WeightedKgBlend surpasses standalone methods, demonstrating superior predictive accuracy when tested to discover drug-disease candidates over a large-scale biomedical knowledge graph called Mechanistic Repositioning Network. This research highlights the efficacy of an integrated approach combining multiple methods in drug discovery, showcasing improved performance and the potential for transformative insights in the realm of biomedical knowledge discovery.
Matching journals
The top 8 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- A Graph-Attention-Based Deep Learning Network for Predicting Biotech-Small-Molecule Drug Interactions 96%
- Mining drug-target interactions from biomedical literature using chemical and gene descriptions-based ensemble transformer model. 96%
- FLONE: fully Lorentz network embedding for inferring novel drug targets 95%
Similar papers in this journal
- Ontology-based expansion of virtual gene panels to improve diagnostic efficiency for rare genetic diseases 95%
- Causal Analysis for Multivariate Integrated Clinical and Environmental Exposures Data 94%
- MelAnalyze: Fact-Checking Melatonin claims using Large Language Models and Natural Language Inference 93%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.