Multi-scale Graphical Models for Spatio-Temporal Processes
Authors: firdaus janoos, Huseyin Denli, Niranjan Subrahmanya
NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate results on a general class of problems arising in exploration geophysics by discovering graphical structure that is physically meaningful and provide evidence of its advantages over alternative approaches. In this section we present an application to determining the connectivity structure of a medium from data of flow through it under a potential/pressure field. |
| Researcher Affiliation | Industry | Firdaus Janoos Huseyin Denli Niranjan Subrahmanya Exxon Mobil Corporate Strategic Research Annandale, NJ 08801 |
| Pseudocode | No | The paper describes optimization steps and algorithms (e.g., Block Coordinate Descent) in text, but it does not include a structured pseudocode block or algorithm listing. |
| Open Source Code | No | The paper states, 'the data and model are available on request'. This does not constitute concrete access to open-source code. |
| Open Datasets | No | The paper states that 'The data were generated by numerical simulation of eqn. (5) using a proprietary high-fidelity solver' and 'the data and model are available on request', indicating it is not a publicly accessible dataset. |
| Dataset Splits | Yes | In all cases, model hyperparameters were selected via 10-fold cross-validation described in Supplemental Appendix G. in terms of misfit (i.e. training ) error and in terms of cross-validation (i.e. testing) error |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used to run the experiments. |
| Software Dependencies | No | The paper mentions 'a proprietary high-fidelity solver' for data generation but does not list specific software dependencies with version numbers (e.g., Python, PyTorch, or specific solvers with versions) used for their model implementation. |
| Experiment Setup | Yes | Estimation was done for K = 20, with multiple initializations and hyper-parameter selection as described above. The K-means step was initialized by distributing seed locations uniformly at random. The model orders P and Q were kept constant at 50 and 25 respectively. |