Near Optimal Best Arm Identification for Clustered Bandits

Authors: Yash, Avishek Ghosh, Nikhil Karamchandani

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on synthetic and real-world (Movie Lens, Yelp) data demonstrates the superior performance of the proposed algorithms in terms of sample and communication efficiency, particularly in settings where M N.
Researcher Affiliation Academia 1Department of Electrical Engineering, IIT Bombay, India 2Department of Computer Science and Engineering, IIT Bombay, India. Correspondence to: Avishek Ghosh <avishek EMAIL>.
Pseudocode Yes Algorithm 1 Cl-BAI Algorithm 2 SE (A,γ,R) Algorithm 3 BAI-Cl Algorithm 4 d SE(S,µS,γ,η,η1)
Open Source Code No The paper describes algorithms and experiments but does not provide any explicit statement about releasing source code, nor does it include a link to a code repository.
Open Datasets Yes Experiments on synthetic and real-world (Movie Lens, Yelp) data... We perform experiments using the Movie Lens-1M dataset... We perform experiments using the Yelp6 dataset, which contains ratings for various businesses given by users across different states of the US. 6https://www.yelp.com/dataset
Dataset Splits No For the three datasets constructed above, we vary N and plot the average number of pulls for the various schemes in Figures 1(a)(b)(c)... Finally, there are N agents, divided into N/M sized clusters... We group the users into six age categories: 18 24, 25 34, 35 44, 45 49, 50 55, and 56+... As before, we assume that there are N agents divided into M equal-sized clusters. The paper describes the creation of clusters and varying parameters like N, but does not specify explicit train/test/validation dataset splits typically used for model evaluation.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory) used to conduct the experiments.
Software Dependencies No The paper describes the algorithms (e.g., Successive Elimination) but does not list any specific software dependencies or libraries with version numbers (e.g., Python, PyTorch, TensorFlow, scikit-learn versions).
Experiment Setup Yes We set the error probability δ =10^-10 for our experiments and present sample complexity results which are averaged over multiple independent runs of the corresponding algorithms... Input: δ, η; Initialize: Best Arm 0N ... γ =( δ 12NK )^4/3,R=log(17/η)... γ =δ/(2M)... Input: η,δ... γ = δ.log( M M 1 ) δ ) ,R=log(1/η)+1)... γ =δ/3N... S,µS,γ,η,η1... δk =10^-k...