Continuous Time Bayesian Networks with Clocks
Authors: Nicolai Engelmann, Dominik Linzner, Heinz Koeppl
ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we give simulations for parameter and structure inference based on synthetic trajectories from augmented CTBNs generated using the Gillespie sampling procedure mentioned in section 2.1. In the end, we highlight a result for structure inference of an augmented CTBN under a Gamma distribution assumption on trajectories obtained through Gene Net Weaver (Schaffter et al., 2011) for a gene regulatory network. |
| Researcher Affiliation | Academia | 1Department of Electrical Engineering and Information Technology, Technische Universitat Darmstadt, Darmstadt, Germany 2Department of Biology, Technische Universitat Darmstadt, Darmstadt, Germany. |
| Pseudocode | Yes | We are now in the position to generate CTBNs with any parametrized distribution for any state xn and condition un for all n. It is no coincidence, that (4) resembles the density of a minimum distribution, which renders trajectory sampling straight forward and efficient for parametric local survival time distributions despite having to evaluate truncated distributions. This can be done with a Gillespie algorithm (outlined in Appendix C), subsequently drawing global survival times, changing processes and next states. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. |
| Open Datasets | Yes | In the end, we highlight a result for structure inference of an augmented CTBN under a Gamma distribution assumption on trajectories obtained through Gene Net Weaver (Schaffter et al., 2011) for a gene regulatory network. |
| 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 | Their shapes and rates (Gamma: α1 40, β1 5 for the shape and α2 25, β2 2.5 for the rate generation, Weibull: α1 8, β1 0.5 for the shape and α2 5, β2 3 for the rate) were chosen to lead to reasonable model parameters with a significant variance in the samples and pronounced non-exponential appearance. |