基于多臂赌博机模型的信道选择

在择机频谱接入系统中,为解决未知信道环境先验知识下的信道选择问题,提出了一种基于多臂赌博机(MAB)模型的改进UCB(Upper Confidence Bound) 索引选择策略。该策略是通过在UCB索引的置信因子中引入收益方差值来调整对未知信道环境的探索过程,以降低探索成本。结合理论证明了本策略有较快的收敛速度,还证明了本策略下的学习后悔值曲线与时隙呈近似对数关系而较缓慢增长。仿真结果表明,与原UCB策略以及贪心算法相比,所提策略更能自适应地选择可用性较好的信道,有效降低学习后悔值并加快其收敛速度,从而提高了系统吞吐量。     

Peasants,bandits, and state intervention: The consolidation of authority in the Ottoman Balkans and Southern Italy

This paper explores the role of bandits and state intervention in the Ottoman Balkans and Southern Italy in the 19th century by using archival documents. I argue that the states may react similarly and radically when their authority is challenged in the periphery. The Ottoman Empire and the Kingdom of Italy developed the same forms of state intervention to fight against the bandits, even though these two states had fundamentally different political, cultural, and socio‐legal structures. I present three different forms of state intervention: (i) victim‐centred state intervention; (ii) security‐centred state intervention; and (iii) authority‐centred state intervention. These three forms consolidated the state's authority while making the two states both fragile and dependent on other social agencies in the long term. I further claim that consolidation of the state's authority manifests the paradox of state intervention and creates more vulnerabilities in traumatic geographies.     

A survey on the bandit problem with switching costs

The paper surveys the literature on the bandit problem, focusing on its recent development in the presence of switching costs. Switching costs between arms makes not only the Gittins index policy suboptimal, but also renders the search for the optimal policy computationally infeasible. This survey will first discuss the decomposability properties of the arms that make the Gittins index policy optimal, and show how these properties break down upon the introduction of costs on switching arms. Having established the failure of the simple index policy, the survey focus on the recent efforts to overcome the difficulty of finding the optimal policy in the bandit problem with switching costs: characterization of the optimal policy, exact derivation of the optimal policy in the restricted environments, and lastly approximation of optimal policy. The advantages and disadvantages of the above approaches are discussed.     

Stopping with congestion and private payoffs

This paper analyses a two-player stopping game with multiarmed bandits in which each player chooses between learning about the quality of her private risky arm and competing for the use of a single shared safe arm. The qualities of the players’ risky arms are independent. A player whose risky arm produces a success no longer competes for the safe arm. We assume that a player observes her opponent’s actions but not his realised payoffs. She is therefore never certain whether her opponent is still competing for the safe arm. When the players’ prior probabilities of success are sufficiently close, there exists no pure strategy equilibrium, and we characterise the unique mixed strategy equilibrium. Otherwise, the unique equilibrium is in pure strategies. The amount of experimentation performed in equilibrium is inefficiently low but, for many priors, higher than if successes are publicly observed.     

Learning Dynamics with Private and Public Signals

This paper studies the dynamics of learning in a model of technology adoption. Firms learn about an unknown technology by observing both private and public signals. Because of the externality associated with the public signal, the social planner has firms experiment more in the initial period of the model, relative to the market equilibrium. In certain cases, this more rapid generation of information results in the planner experimenting less in later periods of the model. In contrast, typical models with public signals result in the planner inducing more experimentation in all periods of the model relative to the market equilibrium. I would like to thank Matt Mitchell and Tom Holmes for their advice and encouragement. I would also like to thank Thor Koeppl, Cyril Monnet, John Stevens, and Jason Cummins, and two anonymous referees for their helpful comments. This paper is the second chapter of my dissertation. The views expressed herein are my own and not necessarily those of the Bureau of Economic Analysis or the US Department of Commerce.     

Time bandits: How they are created, why they are tolerated, and what can be done about them

Organizations today own little slack, and they must very carefully manage their resources. In this article, we describe an omnipresent, yet often ignored resource utilization problem whereby some workers abandon certain responsibilities and use the freed-up time to pursue personal interests such as hobbies and side businesses. In essence, these “time bandits” work part-time in exchange for full-time pay. While bandits are a minority among workers, their negative effects are significant and widespread. Specifically, banditry undermines an organization's mission, morale, and productivity, as well as putting stakeholder support at risk. In an effort to address this problem, we offer insights in three areas. First, we identify key causes of banditry, including supervisors not enforcing performance standards, poorly constructed reward systems, and the failure to recognize individual differences when designing jobs. Second, we describe reasons why banditry is tolerated within organizations, such as supervisors' desire to avoid conflict and their fear of being labeled as hypocrites. Most importantly, we offer a set of techniques that can prevent and reverse banditry. These include carefully defining expectations, intervening quickly when the symptoms of banditry appear, reducing bandits' compensation over time, and designing jobs that capitalize on individuals' varied skills and motivation.