A Parametric Contextual Online Learning Theory of Brokerage

Authors: François Bachoc, Tommaso Cesari, Roberto Colomboni

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

Reproducibility Variable Result LLM Response
Research Type Theoretical We design algorithms for this problem and prove optimal theoretical regret guarantees under various standard assumptions. ... We prove a key structural result (Lemma 2.1) with two crucial consequences. ... We solve this challenging exploration-exploitation dilemma by proposing an algorithm (Algorithm 1) and prove its optimality by showing a Ld T ln T regret upper bound (Theorem 3.1) and a matching (up to a Ld T lower bound (Theorem 3.2).
Researcher Affiliation Academia 1IMT and IUF, Universit e Paul Sabatier, Toulouse, France 2EECS, University of Ottawa, Ottawa, Canada 3DEIB, Politecnico di Milano, Milano, Italy 4Department of CS, Universit a degli Studi di Milano, Milano, Italy. Correspondence to: Franc ois Bachoc <EMAIL>, Tommaso Cesari <EMAIL>, Roberto Colomboni <EMAIL>.
Pseudocode Yes Algorithm 1 ... Algorithm 2
Open Source Code No The paper does not contain any explicit statement about providing open-source code for the described methodology, nor does it provide a link to a code repository.
Open Datasets No The paper focuses on a theoretical online learning problem and does not present experiments that utilize specific, publicly available datasets. It refers to abstract concepts like 'traders valuations' and 'contexts' without providing concrete dataset access information.
Dataset Splits No The paper is theoretical and does not describe experiments involving specific datasets, therefore, no information regarding training/test/validation dataset splits is provided.
Hardware Specification No The paper is theoretical, focusing on algorithm design, proofs, and regret analysis. It does not describe any experimental setup that would require specific hardware, and thus, no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and describes algorithms and proofs without detailing their implementation. Therefore, it does not specify any software dependencies with version numbers needed for replication.
Experiment Setup No The paper is theoretical and focuses on mathematical proofs, algorithms, and regret analysis. It does not include an experimental section detailing hyperparameters, training configurations, or system-level settings for any empirical evaluation.