桥接对抗性与非平稳多臂老虎机

Context Attentive Bandits: Contextual Bandit with Restricted Context

We consider a novel formulation of the multi-armed bandit model, which we call the contextual bandit with restricted context, where only a limited number of features can be accessed by the learner at every iteration. This novel formulation is motivated by different online problems arising in clinical trials, recommender systems and attention modeling. Herein, we adapt the standard multi-armed bandit algorithm known as Thompson Sampling to take advantage of our restricted context setting, and propose two novel algorithms, called the Thompson Sampling with Restricted Context(TSRC) and the Window…

Evolutionary Multi-Armed Bandits with Genetic Thompson Sampling

As two popular schools of machine learning, online learning and evolutionary computations have become two important driving forces behind real-world decision making engines for applications in biomedicine, economics, and engineering fields. Although there are prior work that utilizes bandits to improve evolutionary algorithms' optimization process, it remains a field of blank on how evolutionary approach can help improve the sequential decision making tasks of online learning agents such as the multi-armed bandits. In this work, we propose the Genetic Thompson Sampling, a bandit algorithm that…

多智能体多臂老虎机中的遗憾下界

Regret Lower Bounds in Multi-agent Multi-armed Bandit

本文研究多智能体多臂老虎机的遗憾下界，分别在随机与对抗奖励、连通与非连通图设置下建立紧致下界，并弥合对抗情形中既有上下界的差距。

具有非平稳奖励的多智能体组合赌博机的有限时间保证

Finite-Time Guarantees for Multi-Agent Combinatorial Bandits with Nonstationary Rewards

针对干预效应随时间变化的多智能体序贯资源分配问题，提出含习惯化与恢复机制的非平稳组合多臂赌博机框架，并给出具有动态遗憾保证的算法，糖尿病干预案例显示注册提升最高达基线三倍。

关于在满足满意遗憾最小化中即使少量非平稳性的危险

On the Peril of (Even a Little) Nonstationarity in Satisficing Regret Minimization

研究非平稳K臂老虎机的满足满意遗憾保证，发现当L≥2时最优遗憾为Θ(L log T)，而L=1时可实现T无关的Θ(1)遗憾。

神经索引策略用于具有异构预算的静止多动作臂老虎机

Neural Index Policies for Restless Multi-Action Bandits with Heterogeneous Budgets

针对多动作静止多臂老虎机，提出神经索引策略，通过可微背包层处理异构预算，端到端训练，性能接近最优，可扩展至数百臂。

桥接离线强化学习与模仿学习：一个关于悲观主义的故事

Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism

提出一种统一模仿学习与离线RL的新框架，基于悲观主义的LCB算法能适应未知数据组成，在近专家数据集上实现更快的1/N收敛速率，并在上下文老虎机和MDPs中展现自适应最优性。

On the Hardness of Inventory Management with Censored Demand Data

We consider a repeated newsvendor problem where the inventory manager has no prior information about the demand, and can access only censored/sales data. In analogy to multi-armed bandit problems, the manager needs to simultaneously "explore" and "exploit" with her inventory decisions, in order to minimize the cumulative cost. We make no probabilistic assumptions---importantly, independence or time stationarity---regarding the mechanism that creates the demand sequence. Our goal is to shed light on the hardness of the problem, and to develop policies that perform well with respect to the regre…

GSC-QEMit: 一种基于遥测的分层预测与老虎机框架用于自适应量子误差缓解

GSC-QEMit: A Telemetry-Driven Hierarchical Forecast-and-Bandit Framework for Adaptive Quantum Error Mitigation

本文提出GSC-QEMit框架，通过遥测数据动态切换轻量级抑制与重干预策略，提升逻辑保真度9%并降低不必要的重干预。

在模型不规范情况下的动态决策：一种随机稳定性方法

Dynamic Decision-Making under Model Misspecification: A Stochastic Stability Approach

本文研究了在模型不规范情况下汤普森采样算法的行为和性能，提出了一种统一的随机稳定性框架，为贝叶斯学习与进化动态的结合提供了新的视角。

弥合智能电表数据缺口：面向数据插补的统计模型、机器学习模型与时间序列基础模型基准

Bridging Smart Meter Gaps: A Benchmark of Statistical, Machine Learning and Time Series Foundation Models for Data Imputation

本文构建智能电表缺失数据插补基准，比较两个大语言模型、五个时间序列基础模型及传统机器学习与统计模型，发现时间序列基础模型可在部分场景显著提升精度，但需权衡计算成本。