Temporal Positive-unlabeled Learning for Biomedical Hypothesis Generation via Risk Estimation
Authors: Uchenna Akujuobi, Jun Chen, Mohamed Elhoseiny, Michael Spranger, Xiangliang Zhang
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 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. |
| Researcher Affiliation | Collaboration | 1King Abdullah University of Science and Technology 2 Sony AI, Tokyo |
| Pseudocode | Yes | Algorithm 1: Calculate the future connection score for term pairs ai,j =< vi, vj > |
| Open Source Code | No | The paper does not provide an explicit statement or link to open-source code for the described methodology. |
| Open Datasets | No | The paper describes the construction of its datasets from published papers ('The graphs on which we apply our model are constructed from the title and abstract of papers published in the biomedical fields from 1949 to 2020') but does not provide concrete access (e.g., a URL, DOI, or specific repository name) to these datasets or the scripts to reproduce them. |
| Dataset Splits | No | The paper mentions 'training' and 'testing' data splits, but does not explicitly describe a separate 'validation' split with percentages, counts, or a specific methodology for it. |
| Hardware Specification | Yes | Each GPU based experiment was conducted on an Nvidia 1080TI GPU. |
| Software Dependencies | No | The paper mentions using the 'Tensorflow library' and 'Graph SAGE' but does not specify their version numbers or any other software dependencies with versions. |
| Experiment Setup | Yes | In all our experiments, we set the hidden dimensions to d = 128. For each neural network based model, we performed a grid search over the learning rate lr = {1e 2, 5e 3, 1e 3, 5e 2}. For the prior estimation, we adopted Gaussian, square-root inverse Gamma, and Dirichlet distributions to model the mean, co-variance matrix and mixing coefficient variational posteriors respectively. |