Linear Contextual Bandits With Interference

Authors: Yang Xu, Wenbin Lu, Rui Song

ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The effectiveness of our approach is demonstrated through simulations and synthetic data generated based on Movie Lens data.
Researcher Affiliation Academia 1Department of Statistics, North Carolina State University, USA. Correspondence to: Rui Song <EMAIL>.
Pseudocode Yes Algorithm 1 Linear Contextual Bandits with Interference
Open Source Code Yes All supplementary code is available at our Github repository.
Open Datasets Yes The Movie Lens 1M dataset (Harper & Konstan, 2015) contains over 1 million movie ratings from 6k users, aiding movie recommendations based on historical ratings.
Dataset Splits No Based on the timestamps of each rating and the relative user density, we divided the dataset into T = 200 rounds.
Hardware Specification No All experiments were conducted on a local computer with 16 GB of memory.
Software Dependencies No No specific software dependencies with version numbers are mentioned in the paper.
Experiment Setup Yes The simulation setup of testing coverage probability is as follows. In the estimation of β, the entire process is replicated for B = 1000 times to calculate the empirical coverage. For each replication, we assume there are a total of T = 500 rounds, and we randomly sample the true β from β0 = (2, 3, 1) and β1 = (1, 1, 3) .