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; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]

Distributionally Robust Bayesian Optimization with $\varphi$-divergences

Authors: Hisham Husain, Vu Nguyen, Anton van den Hengel

NeurIPS 2023 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We then show experimentally that our method surpasses existing methods, attesting to the theoretical results. 5 Experiments Experimental setting. The experiments are repeated using 30 independent runs.
Researcher Affiliation Industry Hisham Husain Amazon EMAIL Vu Nguyen Amazon EMAIL Anton van den Hengel Amazon EMAIL
Pseudocode Yes Algorithm 1 DRBO with ฯ•-divergence
Open Source Code No We will release the Python implementation code in the final version.
Open Datasets Yes We consider the popular benchmark functions3 with different dimensions d. ... We perform an experiment on Wind Power dataset [8] and vary the context dimensions |C| {30, 100, 500} in Fig. 4.
Dataset Splits No No specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning was found.
Hardware Specification No No specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) were found for running its experiments.
Software Dependencies No The paper mentions 'Python implementation code' but does not provide specific software dependencies or version numbers (e.g., library names with version numbers).
Experiment Setup Yes Experimental setting. The experiments are repeated using 30 independent runs. We set |C| = 30 which should be suf๏ฌcient to draw c iid q in one-dimensional space to compute Eqs. (4,5). We optimize the GP hyperparameter (e.g., learning rate) by maximizing the GP log marginal likelihood [43].