An algorithm for maximizing target achievement in the stochastic knapsack problem with normal returns

We consider the stochastic linear knapsack problem in which costs are known with certainty but returns are independent, normally distributed random variables. The objective is to maximize the probability that the overall return equals or exceeds a specified target value. A previously proposed preference order dynamic programming-based algorithm has been shown to be potentially suboptimal. We offer an alternative hybrid DP/branch-and-bound algorithm that both guarantees optimality and significantly outperforms generating the set of Pareto optimal returns.© 1993 John Wiley & Sons, Inc.     

Designing rewards-based crowdfunding campaigns for strategic (but distracted) contributors

In rewards-based crowdfunding, entrepreneurs solicit donations from a large number of individual contributors. If total donations exceed a prespecified funding target, the entrepreneur distributes nonmonetary rewards to contributors; otherwise, their donations are refunded. We study how to design such campaigns when contributors choose not just whether to contribute, but also when to contribute. We show that strategic contribution behavior—when contributors intentionally delay until campaign success is likely—can arise from the combination of nonrefundable (potentially very small) hassle costs and the risk of campaign failure, and can explain pledging patterns commonly observed in crowdfunding. Furthermore, such delays do not hurt the entrepreneur if contributors are perfectly rational, but they do if contributors are distracted, that is, if they might fail to return to the campaign after an intentional delay. To mitigate this, we find that an entrepreneur can use a simple menu of rewards with a fixed number of units sold at a low price, and an unlimited number sold at a higher price; this segments contributors over time based on the information they observe upon arrival. We show that, despite its simplicity, such a menu performs well compared to a theoretically optimal menu consisting of an infinite number of different rewards and price levels under many conditions.     

Human Capital,Natural Resource Scarcity and the Rwandan Genocide

Many authors contend that ethnic extremism coupled with political manipulation were the primary factors behind the Rwandan genocide. Yet, to oversimplify the cause of this tragedy makes one blind to the complicated nexus that generated the outcome. Even though this genocide was quick in its execution, the events that lead to this massacre took years to unfold. We argue that the evolution of human capital and the competition for scarce resources contributed to the genocide.     

How not to pay for the war

As a result of allied subsidy and the influence of sunk costs, the marginal cost of the Gulf War to the US was reduced to negligible size. This result is at variance with the Olson‐Zeckhauser thesis that in an alliance “the small exploit the large.” A game theoretic alternative explanation suggests that the relation between allies resembles the game of Chicken, successfully played by the US. The ability to shift the marginal costs of war in the short term raises questions about the possible underestimation of long term effects.     

Forecast horizons for the discounted dynamic lot-size problem allowing speculative motive

This article considers the dynamic lot-size problem under discounting, allowing speculative motive for holding inventory. A variable rolling-horizon procedure is presented, which, under certain regularity conditions, is guaranteed to generate an infinite-horizon optimal-production plan. We also discuss a fixed rolling-horizon procedure which provides a production plan that achieves an infinite-horizon cost within a user-specified tolerance ϵ of optimality. The fixed-horizon length T* needed in this procedure is given in terms of a closed-form formula that is independent of specific forecasted demands. We also present computational results for problems with a range of cost parameters and demand characteristics.     

Optimal block replacement policies with multiple choice at failure

In this article we consider block replacement policies where the operating item is replaced by a new one at times kT, k = 1, 2, …, independently of its failure history. At failure the item is either replaced by a new or a used one or remains inactive until the next planned replacement. The mathematical model is defined and general analytical results are obtained. Computations are carried out for the case where the underlying life distribution is gamma or Weibull.