Thompson Sampling for Multinomial Logit Contextual Bandits

Authors: Min-hwan Oh, Garud Iyengar

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The numerical experiments show that the practical performance of both methods is in line with the theoretical guarantees. and In this section, we perform numerical evaluations to analyze two variants of our proposed algorithm: TS-MNL with optimistic sampling (Algorithm 2) and TS-MNL with the Gaussian approximation for the posterior distribution. We perform both synthetic experiments as well as simulated experiments using a real-world dataset: Movie Lens dataset.
Researcher Affiliation Academia Min-hwan Oh Columbia University New York, NY m.oh@columbia.edu Garud Iyengar Columbia University New York, NY garud@ieor.columbia.edu
Pseudocode Yes Algorithm 1 TS-MNL and Algorithm 2 TS-MNL with Optimistic Sampling
Open Source Code No The paper does not contain any statement about releasing source code or provide a link to a code repository for the methodology described.
Open Datasets Yes We perform both synthetic experiments as well as simulated experiments using a real-world dataset: Movie Lens dataset.2 ... 2https://grouplens.org/datasets/movielens/
Dataset Splits No The paper mentions performing 'synthetic experiments' and 'simulated experiments using a real-world dataset: Movie Lens dataset' and estimates an unknown parameter using 'the entire dataset'. However, it does not explicitly provide specific training, validation, or test dataset split percentages, sample counts, or references to predefined splits within the main text.
Hardware Specification No The paper states 'The details of the experimental setup and additional experimental results are presented in Appendix G.', but no specific hardware details (like GPU/CPU models or memory amounts) are mentioned in the main body of the paper.
Software Dependencies No The paper does not provide specific software dependencies, libraries, or solvers with version numbers that would be needed for replication.
Experiment Setup No The paper mentions 'The details of the experimental setup and additional experimental results are presented in Appendix G.' but does not provide specific experimental setup details, such as concrete hyperparameter values or training configurations, in the main text.