The Reliability of OKRidge Method in Solving Sparse Ridge Regression Problems

Authors: Xiyuan Li, Youjun Wang, Weiwei Liu

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We also conduct experiments to verify our theorems and the results are in excellent agreement with our theoretical findings.
Researcher Affiliation Academia Xiyuan Li Youjun Wang Weiwei Liu School of Computer Science, Wuhan University National Engineering Research Center for Multimedia Software, Wuhan University Institute of Artificial Intelligence, Wuhan University Hubei Key Laboratory of Multimedia and Network Communication Engineering, Wuhan University Lee_xiyuan@outlook.com, youjunw1208@gmail.com, liuweiwei863@gmail.com
Pseudocode No The paper does not contain explicitly labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code No The paper states: 'The codes of the OKRidge algorithm are from [1], which is open.' This indicates that the code for the underlying OKRidge algorithm (developed by others) is open, but not the specific code developed by the authors for their theoretical analysis and experiments.
Open Datasets No The paper describes generating synthetic data for its experiments ('xi is drawn i.i.d. from N(0,I), and ϵi is drawn i.i.d. from N(0, σ2)') but does not use or provide access to a publicly available or open dataset.
Dataset Splits Yes In our experiments, β is randomly generated with β 0 k. For i {1, 2, , n}, xi is drawn i.i.d. from N(0,I), and ϵi is drawn i.i.d. from N(0, σ2). According to the k-sparse linear regression (1), yi = x i β + ϵi, we get dataset (xi, yi) with i = 1, 2, , n. Then, we appy OKRidge to get the estimator ˆβ and calculate the NSE by ˆβ β 2 2 σ2. The NSE is averaged over 10 trials to evaluate the effectiveness of the OKRidge algorithm. In the main paper, we set n d = 0.1, d = 100.
Hardware Specification Yes All experiments were run on the 10x Tensor EX TS2-673917-DPN Intel Xeon Gold 6226 Processor, 2.7Ghz. We set the memory limit to be 100GB.
Software Dependencies No The paper does not specify the versions of any software dependencies used for the experiments (e.g., programming languages, libraries, or solvers).
Experiment Setup Yes In our experiments, β is randomly generated with β 0 k. For i {1, 2, , n}, xi is drawn i.i.d. from N(0,I), and ϵi is drawn i.i.d. from N(0, σ2). According to the k-sparse linear regression (1), yi = x i β + ϵi, we get dataset (xi, yi) with i = 1, 2, , n. Then, we appy OKRidge to get the estimator ˆβ and calculate the NSE by ˆβ β 2 2 σ2. The NSE is averaged over 10 trials to evaluate the effectiveness of the OKRidge algorithm. In the main paper, we set n d = 0.1, d = 100.