Learning Planar Ising Models
Authors: Jason K. Johnson, Diane Oyen, Michael Chertkov, Praneeth Netrapalli
JMLR 2016 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate our method in simulations and for two applications: modeling senate voting records and identifying geo-chemical depth trends from Mars rover data. [...] We present the results of experiments evaluating our algorithm on known models with simulated data to evaluate the correctness of the learned models. |
| Researcher Affiliation | Collaboration | Jason K. Johnson EMAIL Numerica Ft. Collins, CO, USA Diane Oyen EMAIL Michael Chertkov EMAIL Los Alamos National Laboratory Los Alamos, NM, USA Praneeth Netrapalli EMAIL Microsoft Research Cambridge, MA, USA |
| Pseudocode | Yes | Algorithm 1 Greedy Planar Graph Select(P) |
| Open Source Code | No | The paper mentions comparing against a third-party tool, UGMLearn2, and provides its link: "http://www.cs.ubc.ca/~murphyk/Software/L1CRF". However, it does not provide source code for the methodology described in this paper by the authors themselves. |
| Open Datasets | Yes | Our second real-world data set consists of geological observations from the Mars rover Curiosity. [...] NASA data is archived and available at http://pds-geosciences.wustl.edu/missions/msl/chemcam.htm. |
| Dataset Splits | Yes | We run a 10-fold cross-validation, training on 90% of the data and measuring likelihood on the held-out 10% of data. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., CPU/GPU models, memory, or cloud instance types) used for running the experiments. It mentions UGM running for 40 hours without reaching convergence, but this is not a hardware specification. |
| Software Dependencies | No | The paper mentions comparing against "UGMLearn2" and refers to a "Matlab package" but does not provide specific version numbers for any software, including their own implementation or any libraries used. |
| Experiment Setup | Yes | The edge parameters are chosen uniformly randomly between 1 and 1 with the condition that the absolute value be greater than a threshold (chosen to be 0.05) so as to avoid edges with negligible interactions. [...] We solve this unconstrained convex optimization problem using Newton s method with stepsize chosen by back-tracking line search (Boyd and Vandenberghe, 2004). [...] We run a 10-fold cross-validation, training on 90% of the data and measuring likelihood on the held-out 10% of data. |