Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
On Sufficient Graphical Models
Authors: Bing Li, Kyongwon Kim
JMLR 2024 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | By simulation comparisons and an analysis of the DREAM 4 Challenge data set, we demonstrate that our method outperforms the existing methods when the Gaussian or copula Gaussian assumptions are violated, and its performance remains excellent in the high-dimensional setting. (...) In Section 6, we conduct simulation studies to compare of our method with the existing methods. In Section 7, we apply our method to the DREAM 4 Challenge gene network data set. |
| Researcher Affiliation | Academia | Bing Li EMAIL Department of Statistics, Pennsylvania State University 326 Thomas Building, University Park, PA 16802. Kyongwon Kim EMAIL Department of Statistics, Ewha Womans University 52 Ewhayeodae-gil, Seodaemun-gu, Seoul, Republic of Korea, 03760. |
| Pseudocode | Yes | 4.4 Algorithm: Sufficient graphical model |
| Open Source Code | Yes | The code is publicly available on Github: https://github.com/kyongwonkim/Sufficient-Graphical-Model.git. |
| Open Datasets | Yes | By simulation comparisons and an analysis of the DREAM 4 Challenge data set, we demonstrate that our method outperforms the existing methods (...) We now apply sufficient graphical model to a data set from the DREAM 4 Challenge project (...) A description of this data set can be found in Marbach et al. (2010). |
| Dataset Splits | No | The paper describes generating 'n' samples and using '50 samples' for 'averaged ROC curves' for simulation models. For the DREAM 4 Challenge data, it mentions using the dataset but does not specify explicit training/test/validation splits for model training or evaluation in the traditional sense, as the task is network recovery rather than predictive modeling with typical data splits. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU models, CPU types, or memory specifications used for running the experiments. It only mentions general computational aspects without any concrete hardware information. |
| Software Dependencies | No | The paper mentions using the 'Gaussian radial basis function as the kernel' for several methods. It also references 'Kernlab (Karatzoglou et al., 2004)' for parameter choice. However, it does not provide specific version numbers for any software, libraries, or solvers used in the implementation or experimentation. |
| Experiment Setup | Yes | The regularization constants ϵ (i,j) X , ϵ (i,j) X , and ϵ (i,j) U are chosen by the generalized cross validation criterion described in Section 4.3 with the grid {10 ℓ: ℓ= 1, 0, 1, 2, 3, 4}. The kernel parameters γ (i,j) X , γ (i,j) X , γ i,ij XU, γ j,ij XU, and γ ij U are chosen according to (12). (...) The dimension dij for sufficient graphical model is taken to be 2 for all cases (we have also used dij = 1 and the results are very similar to those presented here). |