Contextual Online Bilateral Trade

• URL: http://arxiv.org/abs/2602.12903v1
• Date: Fri, 13 Feb 2026 13:03:10 GMT
• Title: Contextual Online Bilateral Trade
• Authors: Romain Cosson, Federico Fusco, Anupam Gupta, Stefano Leonardi, Renato Paes Leme, Matteo Russo,
• Abstract summary: We study two objectives for this problem, gain from trade and profit.<n>We design algorithms that achieve $O(dlog d)$ regret for gain from trade, and $O(d loglog T + dlog d)$ regret for profit.
• Score: 18.8734045754182
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We study repeated bilateral trade when the valuations of the sellers and the buyers are contextual. More precisely, the agents' valuations are given by the inner product of a context vector with two unknown $d$-dimensional vectors -- one for the buyers and one for the sellers. At each time step $t$, the learner receives a context and posts two prices, one for the seller and one for the buyer, and the trade happens if both agents accept their price. We study two objectives for this problem, gain from trade and profit, proving no-regret with respect to a surprisingly strong benchmark: the best omniscient dynamic strategy. In the natural scenario where the learner observes \emph{separately} whether the agents accept their price -- the so-called \emph{two-bit} feedback -- we design algorithms that achieve $O(d\log d)$ regret for gain from trade, and $O(d \log\log T + d\log d)$ regret for profit maximization. Both results are tight, up to the $\log(d)$ factor, and implement per-step budget balance, meaning that the learner never incurs negative profit. In the less informative \emph{one-bit} feedback model, the learner only observes whether a trade happens or not. For this scenario, we show that the tight two-bit regret regimes are still attainable, at the cost of allowing the learner to possibly incur a small negative profit of order $O(d\log d)$, which is notably independent of the time horizon. As a final set of results, we investigate the combination of one-bit feedback and per-step budget balance. There, we design an algorithm for gain from trade that suffers regret independent of the time horizon, but \emph{exponential} in the dimension $d$. For profit maximization, we maintain this exponential dependence on the dimension, which gets multiplied by a $\log T$ factor.

Related papers

• Nonparametric Contextual Online Bilateral Trade [15.586783656868706]
  We study the problem of contextual online bilateral trade.<n>The goal of the learner is to post prices to facilitate trades between the two parties.<n>We design an algorithm that leverages contextual information through a hierarchical tree construction.
  arXiv  Detail & Related papers  (2026-02-13T13:03:30Z)
• Near-optimal Regret Using Policy Optimization in Online MDPs with Aggregate Bandit Feedback [49.84060509296641]
  We study online finite-horizon Markov Decision Processes with adversarially changing loss and aggregate bandit feedback (a.k.a full-bandit)<n>Under this type of feedback, the agent observes only the total loss incurred over the entire trajectory, rather than the individual losses at each intermediate step within the trajectory.<n>We introduce the first Policy Optimization algorithms for this setting.
  arXiv  Detail & Related papers  (2025-02-06T12:03:24Z)
• Reinforcement Learning with Segment Feedback [56.54271464134885]
  We consider a model named RL with segment feedback, which offers a general paradigm filling the gap between per-state-action feedback and trajectory feedback.<n>Under binary feedback, increasing the number of segments $m$ decreases the regret at an exponential rate; in contrast, surprisingly, under sum feedback, increasing $m$ does not reduce the regret significantly.
  arXiv  Detail & Related papers  (2025-02-03T23:08:42Z)
• Nearly Optimal Algorithms for Contextual Dueling Bandits from Adversarial Feedback [58.66941279460248]
  Learning from human feedback plays an important role in aligning generative models, such as large language models (LLM)<n>We study a model within this domain--contextual dueling bandits with adversarial feedback, where the true preference label can be flipped by an adversary.<n>We propose an algorithm namely robust contextual dueling bandits (RCDB), which is based on uncertainty-weighted maximum likelihood estimation.
  arXiv  Detail & Related papers  (2024-04-16T17:59:55Z)
• No-Regret Learning in Bilateral Trade via Global Budget Balance [29.514323697659613]
  We provide the first no-regret algorithms for adversarial bilateral trade under various feedback models. We show that in the full-feedback model, the learner can guarantee $tilde O(sqrtT)$ regret against the best fixed prices in hindsight. We also provide a learning algorithm guaranteeing a $tilde O(T3/4)$ regret upper bound with one-bit feedback.
  arXiv  Detail & Related papers  (2023-10-18T22:34:32Z)
• An Online Learning Theory of Brokerage [3.8059763597999012]
  We investigate brokerage between traders from an online learning perspective. Unlike other bilateral trade problems already studied, we focus on the case where there are no designated buyer and seller roles. We show that the optimal rate degrades to $sqrtT$ in the first case, and the problem becomes unlearnable in the second.
  arXiv  Detail & Related papers  (2023-10-18T17:01:32Z)
• Repeated Bilateral Trade Against a Smoothed Adversary [5.939280057673226]
  We study repeated bilateral trade where an adaptive $sigma$-smooth adversary generates the valuations of sellers and buyers. We provide a complete characterization of the regret regimes for fixed-price mechanisms under different feedback models.
  arXiv  Detail & Related papers  (2023-02-21T16:30:10Z)
• Horizon-Free and Variance-Dependent Reinforcement Learning for Latent Markov Decision Processes [62.90204655228324]
  We study regret minimization for reinforcement learning (RL) in Latent Markov Decision Processes (LMDPs) with context in hindsight. We design a novel model-based algorithmic framework which can be instantiated with both a model-optimistic and a value-optimistic solver.
  arXiv  Detail & Related papers  (2022-10-20T21:32:01Z)
• An $α$-regret analysis of Adversarial Bilateral Trade [10.275531964940425]
  We study sequential bilateral trade where sellers and buyers valuations are completely arbitrary. We consider gain from trade which is harder to approximate than social welfare.
  arXiv  Detail & Related papers  (2022-10-13T08:57:30Z)
• Minimax-Optimal Multi-Agent RL in Zero-Sum Markov Games With a Generative Model [50.38446482252857]
  Two-player zero-sum Markov games are arguably the most basic setting in multi-agent reinforcement learning. We develop a learning algorithm that learns an $varepsilon$-approximate Markov NE policy using $$ widetildeObigg. We derive a refined regret bound for FTRL that makes explicit the role of variance-type quantities.
  arXiv  Detail & Related papers  (2022-08-22T17:24:55Z)
• The Best of Both Worlds: Reinforcement Learning with Logarithmic Regret and Policy Switches [84.54669549718075]
  We study the problem of regret minimization for episodic Reinforcement Learning (RL) We focus on learning with general function classes and general model classes. We show that a logarithmic regret bound is realizable by algorithms with $O(log T)$ switching cost.
  arXiv  Detail & Related papers  (2022-03-03T02:55:55Z)
• A Deeper Look at Discounting Mismatch in Actor-Critic Algorithms [81.01917016753644]
  We investigate the discounting mismatch in actor-critic algorithm implementations from a representation learning perspective. Theoretically, actor-critic algorithms usually have discounting for both actor and critic, i.e., there is a $gammat$ term in the actor update for the transition observed at time $t$ in a trajectory. Practitioners, however, usually ignore the discounting ($gammat$) for the actor while using a discounted critic.
  arXiv  Detail & Related papers  (2020-10-02T15:51:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.