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].
Improving Sign Random Projections With Additional Information
Authors: Keegan Kang, Weipin Wong
ICML 2018 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We demonstrate the effectiveness of our method on the MNIST test dataset and the Gisette dataset. |
| Researcher Affiliation | Academia | 1Singapore University Of Technology And Design. Correspondence to: Keegan Kang <keegan EMAIL>. |
| Pseudocode | Yes | Algorithm 1 Algorithm For Our Estimator Algorithm 2 General Algorithm For Our Estimators |
| Open Source Code | No | The paper does not provide any concrete access to source code for the methodology, nor does it state that the code is publicly available. |
| Open Datasets | Yes | We run our algorithm on the MNIST test dataset (Lecun et al., 1998) and Gisette dataset (Guyon et al., 2005; Lichman, 2013). The MNIST test dataset has n = 10, 000 observations, and p = 784 parameters. The Gisette dataset has n = 13, 500 observations, and p = 5, 000 parameters. |
| Dataset Splits | No | The paper mentions running experiments on the 'MNIST test dataset' and 'Gisette dataset', but it does not specify any training, validation, or test splits, nor does it reference standard splits with specific details for reproducibility. |
| Hardware Specification | No | The paper does not specify any hardware details such as GPU/CPU models or specific machine configurations used for running experiments. |
| Software Dependencies | No | The paper does not list specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x) that would be needed to replicate the experiments. |
| Experiment Setup | Yes | We set the vector e to be the ο¬rst singular vector of our datasets for reproducibility. We run our experiments for the number of columns k (equivalently number of bits) ranging from {64, 128, . . . , 3008} of our random matrix over 100 simulations for both datasets. |