Reinforcement learning, one of the most active research areas in artificial intelligence, is a computational approach to learning whereby an agent tries to maximize the total amount of reward it receives when interacting with a complex, uncertain environment. In Reinforcement Learning, Richard Sutton and Andrew Barto provide a clear and simple account of the key ideas and algorithms of reinforcement learning. Their discussion ranges from the history of the field's intellectual foundations to the most recent developments and applications. The only necessary mathematical background is familiarity with elementary concepts of probability. The book is divided into three parts. Part I defines the reinforcement learning problem in terms of Markov decision processes. Part II provides basic solution methods: dynamic programming, Monte Carlo methods, and temporal-difference learning. Part III presents a unified view of the solution methods and incorporates artificial neural networks, eligibility traces, and planning; the two final chapters present case studies and consider the future of reinforcement learning.

/pdf/reinforcement-learning-an-introduction-rzxgej9p17.pdf

Reinforcement Learning: An Introduction

The leading and most up-to-date textbook on the far-ranging algorithmic methododogy of Dynamic Programming, which can be used for optimal control, Markovian decision problems, planning and sequential decision making under uncertainty, and discrete/combinatorial optimization. The treatment focuses on basic unifying themes, and conceptual foundations. It illustrates the versatility, power, and generality of the method with many examples and applications from engineering, operations research, and other fields. It also addresses extensively the practical application of the methodology, possibly through the use of approximations, and provides an extensive treatment of the far-reaching methodology of Neuro-Dynamic Programming/Reinforcement Learning.

/pdf/dynamic-programming-and-optimal-control-21dopm2oo5.pdf

Dynamic Programming and Optimal Control

Data Mining - Concepts and Techniques.

/pdf/the-theory-of-industrial-organization-19bejph0ej.pdf

The Theory of Industrial Organization

Analytic hierarchy process : An overview of applications

Electronic Data Interchange EDI is an emerging type of standardized inter-organizational information system. We analyze the impact of EDI on the upstream suppliers' competitive position in a simple two-level hierarchical market structure where the buyer faces a linear demand curve and the competing heterogeneous suppliers have an upward-sloping marginal cost function. We show that a suppliers' adoption of EDI can generate positive externalities for the buyer and negative or competitive externalities for other suppliers. As a result, the buyer provides a price premium to those suppliers who adopt EDI and increases their sales volume and market share. Moreover, when the benefits that the buyer can derive from implementing EDI are substantial, and the suppliers' EDI adoption costs are high, it may be in the buyer's best interest to subsidize the suppliers so as to encourage them to adopt EDI, instead of mandating them to do so. Regardless of whether the buyer employs a mandatory or a subsidizing policy, the buyer and the end consumers may be the only ones who gain from this new technology. Consequently, a partial adoption by the supplier base may be optimal for the buyer when the suppliers' adoption costs are sufficiently high. We also show that, while EDI reduces the transaction costs of the buyer, the upstream market tends to become more concentrated as a result of increased cost differentials. These results provide one economic explanation of the fact that many companies have actually reduced their supplier base after implementing EDI, despite a significant reduction in their market transaction costs.

Electronic data interchange: competitive externalities and strategic implementation policies

Generalized polynomial approximations in Markovian decision processes

The Internet's computational power and flexibility have made auctions a widespread and integral part of both consumer and business markets. Though online auctions are a multi-billion dollar annual activity, with a growing variety of sophisticated trading mechanisms, scientific research on them is at an early stage. This paper analyzes the current state of management science research on online auctions. It develops a broad research agenda for issues such as the behavior of online auction participants, the optimal design of online auctions, the integration of auctions into the ongoing operation of firms, and the use of the data generated by online auctions to inform future trading mechanisms. These research areas will draw from applied and theoretical work spanning management science, economics, and information systems.

http://www.simon.rochester.edu/fac/pinker/Papers/ManagingOnlineAuctions.pdf

Managing Online Auctions: Current Business and Research Issues

Managing Online Auctions:Current Business and Research Issues.: Current Business and Research Issues.

Economic lot sizing and batching models often assume reliable manufacturing facilities. In this research, we focus on the effects of machine breakdowns and corrective maintenance on the economic lot sizing decisions. Two production control policies are proposed for coping with these stochastic interferences. The first policy assumes that production of the interrupted lots is not resumed after a breakdown. Instead, the on-hand inventory is depleted before a new cycle is initiated. Under the second policy studied here, production is immediately resumed after a breakdown, if the current on-hand inventory is below a certain threshold level. It is shown that this control structure is optimal among all stationary policies. We show that under both policies the optimal lot sizes will always be bigger than the ones in a corresponding deterministic case, and that the optimal lot size increases with the failure rate. We also provide exact optimal and closed form approximate lot sizing formulas and derive tight bound...

Abraham Seidmann

Papers

Electronic data interchange: competitive externalities and strategic implementation policies

Generalized polynomial approximations in Markovian decision processes

Managing Online Auctions: Current Business and Research Issues

Managing Online Auctions:Current Business and Research Issues.: Current Business and Research Issues.

Production lot sizing with machine breakdowns