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].
A Unified Convex Surrogate for the Schatten-pNorm
Authors: Chen Xu, Zhouchen Lin, Hongbin Zha
AAAI 2017 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on both synthetic and real datasets exhibit its superior performance over the state-of-the-art methods. Its speed is also highly competitive. |
| Researcher Affiliation | Academia | Chen Xu, Zhouchen Lin, Hongbin Zha Key Laboratory of Machine Perception (MOE), School of EECS, Peking University, P. R. China Cooperative Medianet Innovation Center, Shanghai Jiao Tong University, P. R. China EMAIL, EMAIL, EMAIL |
| Pseudocode | Yes | Algorithm 1 Minimizing F(X) in (16) with accelerated PALM. |
| Open Source Code | No | The paper provides links to third-party code used for comparison, but no explicit statement or link for the authors' own source code for the methodology described. |
| Open Datasets | Yes | We conduct experiments on three real-world recommendation system datasets: Movie Lens 1M, Movie Lens 10M11, and Netflix (SIGKDD 2007). |
| Dataset Splits | Yes | Following the experimental setup in (Shang, Liu, and Cheng 2016a), we randomly pick out 80% of the observed entries as the training data and use the remaining 20% for testing. |
| Hardware Specification | Yes | All the codes are run in Matlab on a desktop PC with a 3.4 GHz CPU and 20 GB RAM. |
| Software Dependencies | No | All the codes are run in Matlab on a desktop PC with a 3.4 GHz CPU and 20 GB RAM. |
| Experiment Setup | Yes | As done in (Lai, Xu, and Yin 2013), d is overestimated as 3 * 5 = 15. ... Here we fix the regularization λ = 200 and tune it for other algorithms in the range [1, 200]. Following the experimental setup in (Shang, Liu, and Cheng 2016a), we randomly pick out 80% of the observed entries as the training data and use the remaining 20% for testing. ... Results (d = 10 for the factorization formulation) of all compared algorithms are shown in the first row of Fig. 3. |