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].

Non-Convex Matrix Completion and Related Problems via Strong Duality

Authors: Maria-Florina Balcan, Yingyu Liang, Zhao Song, David P. Woodruff, Hongyang Zhang

JMLR 2019 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental This work studies the strong duality of non-convex matrix factorization problems... We apply our framework to two prototypical matrix factorization problems: matrix completion and robust Principal Component Analysis... Our framework shows that exact recoverability and strong duality hold with nearly-optimal sample complexity for the two problems. ... In this section, we will present our experimental results. 8.1. Experiments on Synthetic Data 8.2. Experiments on Real Data
Researcher Affiliation Academia Maria-Florina Balcan Carnegie Mellon University, Yingyu Liang University of Wisconsin-Madison, Zhao Song UT-Austin & Harvard University, David P. Woodruff Carnegie Mellon University, Hongyang Zhang Carnegie Mellon University & TTIC
Pseudocode No The paper describes algorithms and methods but does not include any explicitly labeled pseudocode or algorithm blocks. It references algorithms like the Douglas-Rachford algorithm but does not present them as pseudocode within the paper.
Open Source Code No The paper does not provide concrete access to source code for the methodology described. It thanks collaborators for providing code for a specific function, but does not state that the authors' own implementation is publicly available.
Open Datasets Yes To verify the performance of the algorithms on real data, we conduct experiments on the Hopkins 155 data set.
Dataset Splits Yes We uniformly sample m entries from the matrix as our observations and run the matrix completion algorithms. ... m = 0.05n1n2 m = 0.1n1n2
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU, GPU models, memory, or cloud instances) used for running the experiments.
Software Dependencies No The paper mentions using the 'Augmented Lagrange Multiplier Method' and refers to the 'Douglas-Rachford algorithm' but does not specify any software libraries or tools with version numbers.
Experiment Setup Yes We set the parameter r in r minimization (15) as the true rank, and use the Augmented Lagrange Multiplier Method (Chen et al., 2009) for optimization... The parameter r in the r minimization is set as the number of moving objects which is known to us in the data set.