Learning Kernelized Contextual Bandits in a Distributed and Asynchronous Environment

Authors: Chuanhao Li, Huazheng Wang, Mengdi Wang, Hongning Wang

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental To validate Async-Kernel UCB s effectiveness in reducing communication cost, we performed extensive empirical evaluations on both synthetic and real-world datasets, and reported the results (over 10 runs) in Figure 2.
Researcher Affiliation Academia Chuanhao Li1 Huazheng Wang2 Mengdi Wang3 Hongning Wang1 1University of Virginia 2Oregon State University 3Princeton University
Pseudocode Yes Algorithm 1 Asynchronous Kernel UCB (Async-Kernel UCB)
Open Source Code No The paper does not contain any explicit statements or links indicating that source code for the described methodology is publicly available.
Open Datasets Yes UCI Machine Learning Repository (Dua & Graff, 2017) (...) Movie Lens consists of 25 million ratings between 160 thousand users and 60 thousand movies (Harper & Konstan, 2015).
Dataset Splits No The paper does not specify exact percentages or sample counts for training, validation, or test dataset splits. It describes how datasets were partitioned or features extracted, but not the specific data partitioning for model training and evaluation.
Hardware Specification No The paper does not explicitly describe the specific hardware (e.g., GPU/CPU models, memory) used to run the experiments.
Software Dependencies No The paper mentions using a 'Gaussian kernel' and 'Sigmoid function' but does not specify any software libraries or their version numbers (e.g., Python, PyTorch, TensorFlow, scikit-learn) used for implementation or experimentation.
Experiment Setup Yes Synthetic dataset We simulated the distributed bandit setting in Section 3.1, with d = 20, T = 104, N = 102. At each time step t [T], client it [N] selects an arm from candidate set At (with |At| = 20) (...) For all the kernel bandit algorithms, we used the Gaussian kernel k(x, y) = exp( γ x y 2), where we did a grid search of γ {0.1, 1, 4}, and for Fed GLBUCB, we used Sigmoid function µ(z) = (1 + exp( z)) 1 as link function. For all algorithms, instead of using their theoretically derived exploration coefficient α, we followed the convention Li et al. (2010a); Zhou et al. (2020) to use grid search for α in {0.1, 1, 4}.