Faster Randomized Infeasible Interior Point Methods for Tall/Wide Linear Programs
Authors: Agniva Chowdhury, Palma London, Haim Avron, Petros Drineas
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our empirical evaluations verify our theoretical results on both real and synthetic data. |
| Researcher Affiliation | Academia | Agniva Chowdhury Department of Statistics Purdue University West Lafayette, IN, USA chowdhu5@purdue.edu Palma London ORIE Department Cornell University Ithaca, NY, USA plondon@cornell.edu Haim Avron School of Mathematical Sciences Tel Aviv University Tel Aviv, Israel haimav@tauex.tau.ac.il Petros Drineas Department of Computer Science Purdue University West Lafayette, IN, USA pdrineas@purdue.edu |
| Pseudocode | Yes | Algorithm 1 Solving eqn. (5) via CG; Algorithm 2 Infeasible IPM |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. It does not include a link to a repository or explicitly state that the code is released. |
| Open Datasets | Yes | We demonstrate the empirical performance of our algorithm on a variety of synthetic and real-world datasets from the UCI ML Repository [20], such as ARCENE, DEXTER [23], Driv Face [16], and a gene expression cancer RNA-Sequencing dataset that is part of the PANCAN dataset [50]. |
| Dataset Splits | No | The paper references datasets and their use in experiments but does not explicitly provide specific dataset split information (percentages, sample counts, or citations to predefined splits) for training, validation, and testing needed to reproduce the data partitioning. It mentions 'binary classification' and 'number of training points' but no clear splits. |
| Hardware Specification | Yes | The experiments were implemented in Python and run on a server with Intel E52623V3@3.0GHz 8 cores and 64GB RAM. |
| Software Dependencies | No | The paper states, "The experiments were implemented in Python" but does not specify version numbers for Python or any key libraries, solvers, or other software components used. |
| Experiment Setup | Yes | If not specified: τ = 10 9, tol CG = 10 5, and σ = 0.5. We evaluated a Gaussian sketching matrix and the initial triplet (x, y, s) for all IPM algorithms was set to be all ones. |