Variational Network Inference: Strong and Stable with Concrete Support
Authors: Amir Dezfouli, Edwin Bonilla, Richard Nock
ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We show that, on three applications, our approach outperforms previous methods consistently. ... We benchmark our approach against the state of the art on three very different and challenging problems: discovering brain functional connectivity, modeling property prices in Sydney, and understanding regulation in the yeast genome. We provide a quantitative evaluation of our approach, showing that it consistently outperforms competitive baselines. |
| Researcher Affiliation | Collaboration | 1 UNSW, Sydney. 2 Started work at Data61. 3 Data61, the Australian National University and the University of Sydney. |
| Pseudocode | No | No structured pseudocode or algorithm blocks (clearly labeled algorithm sections or code-like formatted procedures) were found in the paper. |
| Open Source Code | No | The paper does not contain an explicit statement about making its source code available or provide a link to a code repository for the methodology described. |
| Open Datasets | Yes | We analyzed the benchmarks of Smith et al. (2011)... (Section 5.1). ...Saccharomyces cerevisiae (Spellman et al., 1998). This represents 100,000+ data points and a network with up to 38,000,000+ arcs. The true underlying network is unknown but there is extensive literature about its major features. Here we take as references the cell cycle transcriptionally regulated genes (Rowicka et al., 2007) and http://www.yeastgenome.org as a more general resource. |
| Dataset Splits | No | The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning. |
| Hardware Specification | No | The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | For more details of the baseline methods, prior setting and optimization specifics see the supplement, III. ... We set the discrimination threshold for each method so that on average each method finds 17-19 edges in the network. |