Case‐based reasoning and improved adaptive search for project scheduling

Most scheduling problems are notoriously intractable, so the majority of algorithms for them are heuristic in nature. Priority rule‐based methods still constitute the most important class of these heuristics. Of these, in turn, parametrized biased random sampling methods have attracted particular interest, due to the fact that they outperform all other priority rule‐based methods known. Yet, even the “best” such algorithms are unable to relate to the full range of instances of a problem: Usually there will exist instances on which other algorithms do better. We maintain that asking for the one best algorithm for a problem may be asking too much. The recently proposed concept of control schemes, which refers to algorithmic schemes allowing to steer parametrized algorithms, opens up ways to refine existing algorithms in this regard and improve their effectiveness considerably. We extend this approach by integrating heuristics and case‐based reasoning (CBR), an approach that has been successfully used in artificial intelligence applications. Using the resource‐constrained project scheduling problem as a vehicle, we describe how to devise such a CBR system, systematically analyzing the effect of several criteria on algorithmic performance. Extensive computational results validate the efficacy of our approach and reveal a performance similar or close to state‐of‐the‐art heuristics. In addition, the analysis undertaken provides new insight into the behaviour of a wide class of scheduling heuristics. © 2000 John Wiley & Sons, Inc. Naval Research Logistics 47: 201–222, 2000     

Selecting T for a periodic review inventory model with staggered deliveries

Consider a single‐item, periodic review, infinite‐horizon, undiscounted, inventory model with stochastic demands, proportional holding and shortage costs, and full backlogging. Orders can arrive in every period, and the cost of receiving them is negligible (as in a JIT setting). Every T periods, one audits the current stock level and decides on deliveries for the next T periods, thus incurring a fixed audit cost and—when one schedules deliveries—a fixed order cost. The problem is to find a review period T and an ordering policy that satisfy the average cost criterion. The current article extends an earlier treatment of this problem, which assumed that the fixed order cost is automatically incurred once every T periods. We characterize an optimal ordering policy when T is fixed, prove that an optimal review period T** exists, and develop a global search algorithm for its computation. We also study the behavior of four approximations to T** based on the assumption that the fixed order cost is incurred during every cycle. Analytic results from a companion article (where μ/σ is large) and extensive computational experiments with normal and gamma demand test problems suggest these approximations and associated heuristic policies perform well when μ/σ ≥ 2. © 2000 John Wiley & Sons, Inc. Naval Research Logistics 47: 329–352, 2000     

A sequential auction‐bargaining procurement model

One of the most important decisions that a firm faces in managing its supply chain is a procurement decision: selecting suitable suppliers among many potential competing sellers and reducing the purchase cost. While both auctions and bargaining have been extensively studied in the literature, the research that combines auctions and bargaining is limited. In this article, we consider a combined auction‐bargaining model in a setting where a single buyer procures an indivisible good from one of many competing sellers. The procurement model that we analyze is a sequential model consisting of the auction phase followed by the bargaining phase. In the auction phase, the sellers submit bids, and the seller with the lowest bid is selected as the winning bidder. In the bargaining phase, the buyer audits the cost of the winning seller and then negotiates with him to determine the final price. For this auction‐bargaining model, we find a symmetric equilibrium bidding strategy for the sellers in a closed form, which is simple to understand and closely related to the classical results in the auction and bargaining literature. We also show that the auction‐bargaining model generates at least as much profit to the buyer as the standard auction or sequential bargaining model. © 2009 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

Concise RLT forms of binary programs: A computational study of the quadratic knapsack problem

The reformulation‐linearization technique (RLT) is a methodology for constructing tight linear programming relaxations of mixed discrete problems. A key construct is the multiplication of “product factors” of the discrete variables with problem constraints to form polynomial restrictions, which are subsequently linearized. For special problem forms, the structure of these linearized constraints tends to suggest that certain classes may be more beneficial than others. We examine the usefulness of subsets of constraints for a family of 0–1 quadratic multidimensional knapsack programs and perform extensive computational tests on a classical special case known as the 0–1 quadratic knapsack problem. We consider RLT forms both with and without these inequalities, and their comparisons with linearizations derived from published methods. Interestingly, the computational results depend in part upon the commercial software used. © 2009 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

Dynamic coordination of production planning and sales admission control in the presence of a spot market

We consider the decision‐making problem of dynamically scheduling the production of a single make‐to stock (MTS) product in connection with the product's concurrent sales in a spot market and a long‐term supply channel. The spot market is run by a business to business (B2B) online exchange, whereas the long‐term channel is established by a structured contract. The product's price in the spot market is exogenous, evolves as a continuous time Markov chain, and affects demand, which arrives sequentially as a Markov‐modulated Poisson process (MMPP). The manufacturer is obliged to fulfill demand in the long‐term channel, but is able to rein in sales in the spot market. This is a significant strategic decision for a manufacturer in entering a favorable contract. The profitability of the contract must be evaluated by optimal performance. The current problem, therefore, arises as a prerequisite to exploring contracting strategies. We reveal that the optimal strategy of coordinating production and sales is structured by the spot price dependent on the base stock and sell‐down thresholds. Moreover, we can exploit the structural properties of the optimal strategy to conceive an efficient algorithm. © 2010 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

Optimal control of a production‐inventory system with both backorders and lost sales

We consider the optimal control of a production inventory‐system with a single product and two customer classes where items are produced one unit at a time. Upon arrival, customer orders can be fulfilled from existing inventory, if there is any, backordered, or rejected. The two classes are differentiated by their backorder and lost sales costs. At each decision epoch, we must determine whether or not to produce an item and if so, whether to use this item to increase inventory or to reduce backlog. At each decision epoch, we must also determine whether or not to satisfy demand from a particular class (should one arise), backorder it, or reject it. In doing so, we must balance inventory holding costs against the costs of backordering and lost sales. We formulate the problem as a Markov decision process and use it to characterize the structure of the optimal policy. We show that the optimal policy can be described by three state‐dependent thresholds: a production base‐stock level and two order‐admission levels, one for each class. The production base‐stock level determines when production takes place and how to allocate items that are produced. This base‐stock level also determines when orders from the class with the lower shortage costs (Class 2) are backordered and not fulfilled from inventory. The order‐admission levels determine when orders should be rejected. We show that the threshold levels are monotonic (either nonincreasing or nondecreasing) in the backorder level of Class 2. We also characterize analytically the sensitivity of these thresholds to the various cost parameters. Using numerical results, we compare the performance of the optimal policy against several heuristics and show that those that do not allow for the possibility of both backordering and rejecting orders can perform poorly.© 2010 Wiley Periodicals, Inc. Naval Research Logistics 2010     

Path optimization for the resource‐constrained searcher

We formulate and solve a discrete‐time path‐optimization problem where a single searcher, operating in a discretized three‐dimensional airspace, looks for a moving target in a finite set of cells. The searcher is constrained by maximum limits on the consumption of one or more resources such as time, fuel, and risk along any path. We develop a specialized branch‐and‐bound algorithm for this problem that uses several network reduction procedures as well as a new bounding technique based on Lagrangian relaxation and network expansion. The resulting algorithm outperforms a state‐of‐the‐art algorithm for solving time‐constrained problems and also is the first algorithm to solve multi‐constrained problems. © 2010 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

Inventory sharing under decentralized preventive transshipments

We consider preventive transshipments between two stores in a decentralized system with two demand subperiods. Replenishment orders are made before the first subperiod, and the stores may make transshipments to one another between the subperiods. We prove that the transshipment decision has a dominant strategy, called a control‐band conserving transfer policy, under which each store chooses a quantity to transship in or out that will keep its second‐subperiod starting inventory level within a range called a control band. We prove that the optimal replenishment policy is a threshold policy in which the threshold depends on the capacity level at the other store. Finally, we prove that there does not exist a transfer price that coordinates the decentralized supply chain. Our research also explains many of the differences between preventive and emergency transshipments, including differences in the optimal transfer policies and the existence or nonexistence of transfer prices that coordinate the system. © 2010 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

Newsvendor bounds and heuristics for serial supply chains with regular and expedited shipping

We study an infinite‐horizon, N‐stage, serial production/inventory system with two transportation modes between stages: regular shipping and expedited shipping. The optimal inventory policy for this system is a top–down echelon base‐stock policy, which can be computed through minimizing 2N nested convex functions recursively (Lawson and Porteus, Oper Res 48 (2000), 878–893). In this article, we first present some structural properties and comparative statics for the parameters of the optimal inventory policies, we then derive simple, newsvendor‐type lower and upper bounds for the optimal control parameters. These results are used to develop near optimal heuristic solutions for the echelon base‐stock policies. Numerical studies show that the heuristic performs well. © 2009 Wiley Periodicals, Inc. Naval Research Logistics, 2010     

判决特征选择的快速人脸识别算法

为了在确保识别精度的条件下提高人脸识别的速度,提出了判决特征选择算法(SVM-DFS).针对多类分类问题,判决特征选择算法根据统计学习理论使用支持向量机来实现特征选择,根据全概率定理把特征选择和多类分类集成到一个统一框架.在UMIST和FERET人脸数据库上的实验表明:SVM-DFS算法可以用来挑选对分类最有用的特征,这些挑选出来的特征具有明显的物理意义.使用判决特征选择方法不但可以加快分类器的响应速度,而且不降低分类器的泛化能力.