Multiple-Play Bandits in the Position-Based Model

Authors: Paul Lagrée, Claire Vernade, Olivier Cappe

NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In the last section dedicated to experiments, those policies are compared to several benchmarks on both synthetic and realistic data.
Researcher Affiliation Academia Paul Lagrée LRI, Université Paris Sud Université Paris Saclay paul.lagree@u-psud.fr Claire Vernade LTCI, CNRS, Télécom Paris Tech Université Paris Saclay vernade@enst.fr Olivier Cappé LTCI, CNRS Télécom Paris Tech Université Paris Saclay
Pseudocode Yes Algorithm 1 PBM-PIE
Open Source Code No The paper does not provide concrete access to source code for the methodology described. It references a dataset but not its own code.
Open Datasets Yes The dataset was provided for KDD Cup 2012 track 2[1] and involves session logs of soso.com, a search engine owned by Tencent.
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning into train/validation/test sets.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment.
Experiment Setup Yes In order to evaluate our strategies, a simple problem is considered in which K = 5, L = 3, κ = (0.9, 0.6, 0.3) and θ = (0.45, 0.35, 0.25, 0.15, 0.05). ... We conducted a series of 2,000 simulations over this dataset. At the beginning of each run, a query was randomly selected together with corresponding probabilities of scanning positions and arm expectations.