Minimax Rate for Learning From Pairwise Comparisons in the BTL Model

Authors: Julien Hendrickx, Alex Olshevsky, Venkatesh Saligrama

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We perform a number of experiments designed to gauge the accuracy of the WLSM relative to competing methods. Since we are not aware of any real data sets involving comparisons where the true weights are known, we will use synthetic data.
Researcher Affiliation Academia 1ICTEAM, UCLouvain 2Boston University. Correspondence to: Alex Olshevsky <alexols@bu.edu>.
Pseudocode Yes Algorithm 1 Weighted Least Squares Method
Open Source Code No No explicit statement about making the source code available or providing a link to a code repository was found in the paper.
Open Datasets No No explicit access information (link, DOI, formal citation) for a publicly available or open dataset used for training was provided. The paper states, 'we will use synthetic data'.
Dataset Splits No No specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) for validation was provided. The paper mentions using synthetic data for simulations.
Hardware Specification No No specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments were provided in the paper.
Software Dependencies No No specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment were provided in the paper.
Experiment Setup Yes Representative results are shown in Figure 1 for the 2D grid, the 3D grid, and the Erdos Renyi random graph... The 2D and the E-R graph have 100 nodes, while the 3D grid has 125 nodes; the average degree of the E-R graph is 10. Each data point is the average of 50 simulations.