Nonparametric Neighborhood Selection in Graphical Models
Authors: Hao Dong, Yuedong Wang
JMLR 2022 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Simulations indicate that the proposed methods perform well under Gaussian and non-Gaussian settings. We illustrate the proposed methods using two real data examples. |
| Researcher Affiliation | Academia | Hao Dong hao EMAIL Department of Statistics and Applied Probability University of California, Santa Barbara Santa Barbara, CA, USA. Yuedong Wang EMAIL Department of Statistics and Applied Probability University of California, Santa Barbara Santa Barbara, CA, USA. |
| Pseudocode | Yes | Algorithm 1 Input: Data frame X containing n observations with p dimensions. Output: Estimated c, d, θ2, and the neighborhood set nb G(α). |
| Open Source Code | No | The paper describes using existing R packages and modified functions (e.g., 'sscden1 function in the gss package', 'solve.QP in the quadprog package') for their implementation, and mentions using 'author s R codes' for a comparative method (CEF) from another paper. However, it does not provide explicit source code for the methodology developed in this paper, nor a link to a repository for their own implementation. |
| Open Datasets | Yes | The dataset was introduced in Wille et al. (2004) and was downloaded at https://www.ncbi.nlm.nih.gov/pmc/articles/ PMC545783/. |
| Dataset Splits | No | The paper mentions using a '5-fold cross-validation method in all simulations' for selecting the tuning parameter M. It also specifies sample sizes (n=150 and n=300) and repetitions for simulations (100 times). However, it does not provide explicit training/test/validation splits for the primary experimental results or for the real-world applications. |
| Hardware Specification | No | The paper does not provide any specific hardware details such as GPU or CPU models, processor types, or memory used for running the experiments or simulations. |
| Software Dependencies | No | The paper mentions several R packages used for implementation or comparison (e.g., 'R package space', 'R package QUIC', 'R package huge', 'R package spacejam', 'gss package', 'quadprog package'). However, specific version numbers for these software components are not provided. |
| Experiment Setup | Yes | We set dimension p = 20 and consider two sample sizes n = 150 and n = 300. All simulations are repeated for 100 times. We select the tuning parameter M using the 5-fold cross-validation method in all simulations. We set ε = 0.001 in simulation and real data examples. |