Notice: This website is continuously updated as new conference proceedings and journal volumes are published. Consequently, the figures presented here may differ from those reported in our publications.
A brief overview of our methodology is provided on the Methods page, along with references to publications documenting our approach and links to the code and data used in each study.
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.
Adaptive Proximal Average Approximation for Composite Convex Minimization
Authors: Li Shen, Wei Liu, Junzhou Huang, Yu-Gang Jiang, Shiqian Ma
AAAI 2017 | Venue PDF | LLM Run Details | Input Tokens: 14,267 Total number of tokens sent to the LLM as input for this paper's analysis. | Output Tokens: 6,029 Total number of tokens produced by the LLM (including reasoning/thinking tokens) for this paper's analysis.
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section, we perform some experiments on two important problems in machine learning: overlapping group Lasso and graph-guided fused Lasso to verify the efficacy of our proposed APA-APG1 and APA-APG2 algorithms. Figures 1-6 show the performance of PA-APG, APA-APG1, and APA-APG2 on the overlapping group Lasso problem... |
| Researcher Affiliation | Collaboration | Li Shen, Wei Liu, Junzhou Huang, Yu-Gang Jiang, Shiqian Ma School of Mathematics, South China University of Technology, Guangzhou, China Tencent AI Lab, Shenzhen, China Department of Computer Science and Engineering, University of Texas at Arlington, USA School of Computer Science, Fudan University, Shanghai, China Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong, China |
| Pseudocode | Yes | Algorithm 1 APA-APG1 Algorithm Parameters: Choose γ1 > 0, a > 0, and an initial point x0. Let x0 = x0. for k = 0, 1, do Set τk = 1 k+a and γk+1 = min( γ1a k+a, 1 Lf ); xk := (1 τk)xk + τk xk; xk+1 := N i=1 αi Proxγk+1 gγk+1 ( xk γk+1 f( xk)); xk+1 := xk + τ 1 k (xk+1 xk); end for |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described in this paper. |
| Open Datasets | Yes | overlapping group Lasso (Zhao, Rocha, and Yu 2009; Jacob, Obozinski, and Vert 2009; Mairal et al. 2010); graph-guided fused Lasso (Chen et al. 2012; Kim and Xing 2009) |
| Dataset Splits | No | The paper describes data generation parameters (e.g., n=4000, K values, d values) but does not provide specific training/test/validation dataset split information. |
| Hardware Specification | No | The paper does not provide specific hardware details (like GPU/CPU models or memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (like library names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | In the experiments, the sampling dimension n is fixed to n = 4000 and the numbers of groups are set to K = 10, K = 20, and K = 40, respectively. Figures 1-3 show the performance of PA-APG, APA-APG1, and APA-APG2 on the overlapping group Lasso problem with increasing precision parameters ϵ = 1.0e 4, ϵ = 1.0e 5, and ϵ = 1.0e 6, respectively. |