In many industries, managers face the problem of selling a given stock items by deadline. We investigate dynamically pricing such inventories when demand is price sensitive and stochastic firm's objective to maximize expected revenues. Examples that fit this framework include retailers fashion seasonal goods travel leisure industry, which markets space as seats on airline flights, cabins vacation cruises, rooms in hotels become worthless if not sold specific time. formulate using intensity...

Customer choice behavior, such as buy-up and buy-down, is an important phenomenon in a wide range of revenue management contexts. Yet most methodologies ignore this phenomenon—or at best approximate it heuristic way. In paper, we provide exact quite general analysis problem. Specifically, analyze single-leg reserve problem which the buyers' behavior modeled explicitly. The model very general, simply specifying probability purchase for each fare product function set products offered. control...

A firm has inventories of a set components that are used to produce products. There is finite horizon over which the can sell its Demand for each product stochastic point process with an intensity function vector prices products and time at these offered. The problem price finished so as maximize total expected revenue sales horizon. An upper bound on optimal established by analyzing deterministic version problem. solution suggests two heuristics shown be asymptotically volume tends...

Consider a category of product variants distinguished by some attribute such as color or flavor. A retailer must construct an assortment for the category, i.e., select subset to stock and determine purchase quantities each offered variant. We analyze this problem using multinomial logit model describe consumer choice process newsboy represent retailer's inventory cost. show that optimal has simple structure provide insights on how various factors affect level variety. also develop formal...

We propose and analyze a generic mathematical model for dynamic, stochastic vehicle routing problems, the dynamic traveling repairman problem (DTRP). The is motivated by applications in which objective to minimize wait service dynamically changing environment. This departure from classical problems where one seeks total travel time static, deterministic Potential areas of application include repair, inventory, emergency scheduling problems. DTRP defined as follows: Demands arrive according...

We analyze a single-period, stochastic inventory model (newsboy-like model) in which sequence of heterogeneous customers dynamically substitute among product variants within retail assortment when is depleted. The customer choice decisions are based on natural and classical utility maximization criterion. Faced with such substitution behavior, the retailer must choose initial levels for to maximize expected profits. Using sample path analysis, we structural properties profit function. show...

Bid-prices are becoming an increasingly popular method for controlling the sale of inventory in revenue management applications. In this form control, threshold—or “bid”—prices set resources or units (seats on flight legs, hotel rooms specific dates, etc.) and a product (a seat fare class itinerary room sequence dates) is sold only if offered exceeds sum threshold prices all needed to supply product. This approach appealing intuitive practical grounds, but theory underlying it not well...

Gallego et al. [Gallego, G., G. Iyengar, R. Phillips, A. Dubey. 2004. Managing flexible products on a network. CORC Technical Report TR-2004-01, Department of Industrial Engineering and Operations Research, Columbia University, New York.] recently proposed choice-based deterministic linear programming model (CDLP) for network revenue management (RM) that parallels the widely used (DLP) model. While they focused analyzing “flexible products”—a situation in which provider has flexibility using...

We propose a method for estimating substitute and lost demand when only sales product availability data are observable, not all products displayed in periods (e.g., due to stockouts or controls), the seller knows its aggregate market share. The model combines multinomial logit (MNL) choice with nonhomogeneous Poisson of arrivals over multiple periods. Our key idea is view problem terms primary (or first-choice) demand; that is, would have been observed if had available then apply...

In 1991, D. J. Bertsimas and G. van Ryzin introduced analyzed a model for stochastic dynamic vehicle routing in which single, uncapacitated traveling at constant velocity Euclidean region must service demands whose time of arrival, location on-site are stochastic. The objective is to find policy over an infinite horizon that minimizes the expected system (wait plus service) demands. This paper extends our analysis several directions. First, we analyze problem m identical vehicles with...

We analyze a model of inventory competition among n firms that provide competing, substitutable goods. Each firm chooses initial levels for their good in single period (newsboy-like) model. Customers choose dynamically based on current availability, so the at one affect demand all competing firms. This creates strategic interaction firms' decisions. Our work extends earlier variations this problem by Karjalainen (1992), Lippman and McCardle (1997) Parlar (1988). Specifically, we more...

Discrete choice models are appealing for airline revenue management (RM) because they offer a means to profitably exploit preferences attributes such as time of day, routing, brand, and price. They also good at modeling demand unrestricted fare class structures, which widespread throughout the industry. However, there is little empirical research on practicality effectiveness choice-based RM models. Toward this end, we report results study conducted with major U.S. airline. Our had two main...

This paper considers an overbooking problem with multiple reservation and inventory classes, in which the classes may be used as substitutes to satisfy demand of a given class (perhaps at cost). The is jointly determine levels for taking into account substitution options. Such problems arise variety revenue management contexts, including multicabin aircraft, back-to-back scheduled flights on same leg, hotels room types, mixed-vehicle car rental fleets. We model this two-period optimization...

We propose an approach for estimating customer preferences a set of substitutable products using only sales transactions and product availability data. The underlying demand framework combines general, nonparametric discrete choice model with Bernoulli process arrivals over time. is defined by probability mass function (pmf) on possible preference rankings alternatives, it compatible any random utility model. An arriving assumed to purchase the available option that ranks highest in her...

We analyze a randomized version of the deterministic linear programming (DLP) method for computing network bid prices. The consists simulating sequence realizations itinerary demand and solving programs to allocate capacity itineraries each realization. dual prices from this are then averaged form price approximation. This (RLP) is only slightly more complicated implement than DLP method. show that RLP can be viewed as procedure estimating gradient expected perfect information (PI) revenue....

We analyze a dynamic auction, in which seller with C units to sell faces sequence of buyers separated into T time periods. Each group has independent, private values for single unit. Buyers compete directly against each other within period, as traditional and indirectly periods through the opportunity cost capacity assessed by seller. The number well individual buyers' valuations, are random. model is variation leg, multiperiod revenue management problem, consumers act strategically bid...

We investigate a simple adaptive approach to optimizing seat protection levels in airline revenue management systems. The uses only historical observations of the relative frequencies certain seat-filling events guide direct adjustments accordance with optimality conditions Brumelle and McGill (1993). Stochastic approximation theory is used prove convergence this algorithm optimal levels. In simulation study, we compare performance more traditional method that combines censored forecasting...

We consider a revenue management, network capacity control problem in setting where heterogeneous customers choose among the various products offered by firm (e.g., different flight times, fare classes, and/or routings). Customers may therefore substitute if their preferred are not offered. These individual customer choice decisions modeled as very general stochastic sequence of customers, each whom has an ordered list preferences. Minimal assumptions made about statistical properties this...

Virtual nesting is a popular capacity control strategy in network revenue management. In virtual nesting, products (itinerary-fare-class combinations) are mapped (“indexed”) into relatively small number of “virtual classes” on each resource (flight leg) the network. Nested protection levels then used to availability these classes; specifically, product request accepted if and only its corresponding class available required. Bertsimas de Boer proposed an innovative simulation-based...

Consider a firm that sells products over repeated seasons, each of which includes full-price period and markdown period. The may deliberately understock in the to induce high-value customers purchase early at full price. Customers cannot perfectly anticipate availability. Instead, they use observed past capacities form capacity expectations according heuristic smoothing rule. Based on their capacity, decide buy either or We embed this customer learning process dynamic program firm's choices...

In many markets, it is common for headquarters to create a price list but grant local salespeople discretion negotiate prices individual transactions. How much (if any) pricing should topic of debate within firms. We investigate this issue using unique data set from an indirect lender with discretion. estimate that the sales force adjusted in way improved profits by approximately 11% on average. A counterfactual analysis shows centralized, data-driven optimization system could improve even...

In this article we argue that advances made in automation, communication, and manufacturing portend a dramatic reversal of the "bigger is better" approach to cost reductions prevalent many basic infrastructure industries; for example, transportation, electric power generation, raw material processing. We show traditional capital costs achieved by scaling up size are generally matched learning effects mass production process when numbers instead. addition, using U.S. electricity generation...

We propose an expectation-maximization (EM) method to estimate customer preferences for a category of products using only sales transaction and product availability data. The demand model combines general, rank-based discrete choice with Bernoulli process arrivals over time. is defined by probability mass function (pmf) on given set preference rankings alternatives, including the no-purchase alternative. Each represented list, when faced assumed either purchase available option that ranks...

We analyze a class of stochastic and dynamic vehicle routing problems in which demands arrive randomly over time the objective is minimizing waiting time. In our previous work ([6], [7]), we analyzed this problem for case uniformly distributed demand locations Poisson arrivals. paper, using quite different techniques, are able to extend results more realistic where have an arbitrary continuous distribution arrivals follow only general renewal process. Further, improve significantly best...

Customer choice behavior, such as "buy-up" and "buy-down", is an important phenomenon in a wide range of revenue management contexts.Yet most methodologies ignore this -or at best approximate it heuristic way.In paper, we provide exact quite general analysis problem.Specifically, analyze single-leg yield problem which the buyers' behavior modeled explicitly.The model perfectly simply specifies probability purchasing each fare product function set products offered.The control to decide subset...