Temporal Positive-unlabeled Learning for Biomedical Hypothesis Generation via Risk Estimation
This work addresses the challenge of capturing temporal dynamics in biomedical hypothesis generation for researchers, though it is incremental as it builds on existing PU learning methods.
The paper tackled the problem of predicting future connections in dynamic biomedical graphs by formulating hypothesis generation as a positive-unlabeled learning task, and proposed a variational inference model that achieved validated effectiveness on real-world datasets including COVID-19.
Understanding the relationships between biomedical terms like viruses, drugs, and symptoms is essential in the fight against diseases. Many attempts have been made to introduce the use of machine learning to the scientific process of hypothesis generation(HG), which refers to the discovery of meaningful implicit connections between biomedical terms. However, most existing methods fail to truly capture the temporal dynamics of scientific term relations and also assume unobserved connections to be irrelevant (i.e., in a positive-negative (PN) learning setting). To break these limits, we formulate this HG problem as future connectivity prediction task on a dynamic attributed graph via positive-unlabeled (PU) learning. Then, the key is to capture the temporal evolution of node pair (term pair) relations from just the positive and unlabeled data. We propose a variational inference model to estimate the positive prior, and incorporate it in the learning of node pair embeddings, which are then used for link prediction. Experiment results on real-world biomedical term relationship datasets and case study analyses on a COVID-19 dataset validate the effectiveness of the proposed model.