Online metric algorithms with untrusted predictions

Authors: Antonios Antoniadis, Christian Coester, Marek Elias, Adam Polak, Bertrand Simon

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we present an empirical evaluation of our methods on real world datasets, which suggests practicality. (Abstract); 5. Experiments We evaluate the practicality of our approach on real-world datasets for two MTS: caching and ice cream problem. (Section 5)
Researcher Affiliation Academia 1Saarland University and Max-Planck Institute for Informatics, Saarbr ucken, Germany 2CWI, Amsterdam, Netherlands 3EPFL, Lausanne, Switzerland 4Faculty of Mathematics and Computer Science, Jagiellonian University, Krak ow, Poland 5University of Bremen, Bremen, Germany.
Pseudocode Yes We provide a pseudocode in the supplementary material. (Footnote 2, Section 4); Algorithm 1: MIN det (Fiat et al., 1994) and Algorithm 2: MIN rand (Blum & Burch, 2000) are present in Section 2.1.
Open Source Code Yes The source code and datasets are available at Git Hub5. https://github.com/adampolak/ mts-with-predictions (Section 5)
Open Datasets Yes The source code and datasets are available at Git Hub5. https://github.com/adampolak/ mts-with-predictions (Section 5); BK dataset comes from a former social network Bright Kite (Cho et al., 2011). (Section 5.1); Citi dataset comes from a bike sharing platform Citi Bike. (Section 5.1)
Dataset Splits No The paper describes the datasets used (BK, Citi) but does not provide specific details on training, validation, and test splits (e.g., percentages, sample counts, or explicit splitting methodology).
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., CPU, GPU models, memory, or cloud instances) used for running the experiments.
Software Dependencies No The paper does not specify any software dependencies with version numbers (e.g., programming languages, libraries, or specific solvers with their versions) that would be needed for replication.
Experiment Setup Yes We set the parameters to γ = 1 + 0.01 and ϵ = 0.01. These values, chosen from {10 i : i = 0, . . . , 4}, happen to be consistently the best choice in all our experimental settings. (Section 5.1, 'Algorithms' paragraph)