Nuclear Norm Minimization via Active Subspace Selection
Authors: Cho-Jui Hsieh, Peder Olsen
ICML 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We apply our methods to matrix completion problems including Netflix dataset, and show that they are more than 6 times faster than stateof-the-art nuclear norm solvers. |
| Researcher Affiliation | Collaboration | Cho-Jui Hsieh CJHSIEH@CS.UTEXAS.EDU Department of Computer Science, The University of Texas, Austin, TX 78721, USA Peder A. Olsen PEDERAO@US.IBM.COM IBM T.J. Watson Research Center, Yorktown Heights, NY 10598, USA |
| Pseudocode | Yes | Algorithm 1: Our proposed framework |
| Open Source Code | No | The paper does not explicitly provide a link to its source code or state that it is publicly available. While it mentions 'The implementation detail for the competing algorithms are described in Appendix 6.7', this refers to other methods, not the authors' own code for their proposed method. |
| Open Datasets | Yes | We use the real-world recommendation system datasets, Movie Lens (ml100k, ml10m), Netflix, and Yahoo Music (Dror et al., 2012) as shown in Table 1. |
| Dataset Splits | Yes | The balancing parameter λ was chosen by 3-fold cross validation on subsamples and the resulting rank k are also shown in Table 1. |
| Hardware Specification | Yes | All the experiments are conducted on an Intel Xeon X5355 2.66GHz CPU with 32G RAM. |
| Software Dependencies | No | The paper does not specify particular software dependencies with version numbers, such as programming languages, libraries, or frameworks used for implementation. |
| Experiment Setup | Yes | The balancing parameter λ was chosen by 3-fold cross validation on subsamples and the resulting rank k are also shown in Table 1. [...] We set λ = 5, which gives us a solution with rank 186. [...] We solve the nuclear norm regularized multivariate regression problem with λ = 600 to get a solution X with rank 189. |