Bayesian Learning of Dynamic Multilayer Networks
Authors: Daniele Durante, Nabanita Mukherjee, Rebecca C. Steorts
JMLR 2017 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our methods are tested on simulations and infection studies monitoring dynamic face-to-face contacts among individuals in multiple days, where we perform better than current methods in inference and prediction. |
| Researcher Affiliation | Academia | Daniele Durante EMAIL Department of Statistical Sciences University of Padova Padova, 35121, Italy; Nabanita Mukherjee EMAIL Department of Statistical Science Duke University Durham, NC 27708-0251, USA; Rebecca C. Steorts EMAIL Departments of Statistical Science and Computer Science Duke University Durham, NC 27708-0251, USA |
| Pseudocode | Yes | Appendix B. Pseudocode for Posterior Computation. Algorithm 1 provides guidelines for step-by-step implementation of our Gibbs sampler. |
| Open Source Code | No | The paper does not contain any explicit statement about the authors providing open-source code for the methodology described, nor does it provide a link to a code repository. |
| Open Datasets | Yes | Data are available from the human sensing platform Socio Patterns (http://www.sociopatterns.org) and have been collected using wearable devices that exchange low-power radio packets when two individuals are located within a sufficiently close distance to generate a potential occasion of contagion. |
| Dataset Splits | Yes | To assess predictive performance, we perform posterior analysis under our model and the competing methods, holding out from the observed data the networks from time t13 to t17 in the second day. In the application, we hold out the contact network at the last time in the third day Y (3) t14 to assess out-of-sample predictive performance. |
| 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 mentions specific statistical methods and algorithms (e.g., P olya-gamma data augmentation, Gibbs sampler) but does not provide specific ancillary software details like library names with version numbers or programming language versions. |
| Experiment Setup | Yes | We perform posterior computation under our model with κµ = κ x = κx = 0.05 to favor smooth trajectories a priori and a1 = 2, a2 = 2.5 to facilitate adaptation of the latent spaces dimensions. We consider 5000 Gibbs iterations with R = H = 5, and set a burn-in of 1000. |