IMO^3: Interactive Multi-Objective Off-Policy Optimization

Authors: Nan Wang, Hongning Wang, Maryam Karimzadehgan, Branislav Kveton, Craig Boutilier

IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate its effectiveness empirically on several multi-objective optimization problems.
Researcher Affiliation Collaboration Nan Wang1 , Hongning Wang1 , Maryam Karimzadehgan2 , Branislav Kveton3 , Craig Boutilier2 1University of Virginia 2Google Research 3Amazon
Pseudocode Yes Algorithm 1 IMO3
Open Source Code No The paper provides a link to an extended version on arXiv (https://arxiv.org/abs/2201.09798), but does not state that source code for the methodology is openly available or provide a direct link to a code repository.
Open Datasets Yes ZDT1. The ZDT test suite [Zitzler et al., 2000] is the most widely employed benchmark for MOO. We use ZDT1, the first problem in the test suite [...] Yahoo! News Recommendation. This is a news article recommendation problem derived from the Yahoo! Today Module click log dataset (R6A).
Dataset Splits No The paper mentions generating logged data and using it for off-policy evaluation, but it does not provide specific details on train/validation/test splits, percentages, or sample counts for the datasets used in its experiments.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory) used to conduct the experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers (e.g., programming languages, libraries, frameworks).
Experiment Setup No The paper describes the multi-objective problems and baselines, and mentions parameters like 'pre-selection budget L = 500' and 'fixed interaction budget T = 100', but it does not specify concrete hyperparameter values for the models, such as learning rates, batch sizes, or optimizer settings used in training or optimization.