Innovative Applications of O.R.Network revenue management with inventory-sensitive bid prices and customer choice
Highlights
► Solution method for network revenue management problems with improved accuracy compared to other methods. ► Approximate dynamic programming approach with an arbitrary aggregation of inventory units. ► The algorithm allows a trade-off between solution quality and runtime.
Introduction
A particular area of revenue management (RM) that currently receives much interest is the approximate solution of the RM network problem including models of customer choice behavior. Network problems arise in many applications such as hospitality or transportation where the managed products might require more than one resource, for example a hotel that sells rooms over several nights. While network models have been around for some time already, only in recent years researchers devoted themselves to advancing discrete choice models where the purchase decisions also depend on the offered product alternatives. The need for such models is heightened by the rise of low cost service providers since they cut many of the traditional restrictions meant to segment the market, leaving the customer with similar products whose essentially only distinguishing feature is the price. Even if there are still some restrictions, customers increasingly tend to ignore them in their purchase decision so that in some business applications demand can only be observed for the product with the lowest available price, as pointed out by Boyd and Kallesen (2004). Such a behavior is in stark contrast to the traditional independent demand setting where it is assumed that demand is associated with a product and does not depend on market conditions such as which other products the firm offers. Therefore it is crucial to incorporate customer choice models into RM; more on the advantages of customer choice in the RM context can be found in van Ryzin (2005) and, for a comprehensive treatment of RM, in Talluri and van Ryzin (2004b).
We base our investigations on the particularly interesting work of Zhang and Adelman (2009) who extend the previous independent demand RM model of Adelman (2007) to incorporate customer choice behavior. Their approach differs from others in that they use an affine function of the state vector to approximate the value function of the exact dynamic programming formulation with a linear program (LP) in a way such that it yields time-dependent estimates of the marginal capacity values. The optimal objective of this LP constitutes an upper bound on the exact optimal expected revenue which is tighter than those obtained by several other currently available methods. Since the LP possesses many variables, solving the problem by column generation is shown for the multinomial logit choice model (MNL) with disjoint consideration sets to reduce essentially to solving smaller mixed integer linear programs and is thus implementable in practice. They construct policies directly from the dual solution as well as through a dynamic programming decomposition scheme and show that both perform very well. The most important reason for the improved performance is that the LP naturally generates time-dependent marginal capacity value estimates which gives this approach a cutting edge compared to methods that generate static values.
However, intuitively these values should not only depend on time to departure (for the ease of presentation we will stick to airline terminology), but also on the inventory levels. This dependence on intermediate capacity levels of the resources is not captured by current approaches to network RM with choice behavior. In the independent demand setting, a suitable approximation function was recently proposed by Farias and Van Roy (2007). Instead of using constraint generation to deal with the many constraints of the arising linear program they propose using a constraint sampling procedure which is based on the work of de Farias and Van Roy, 2003, de Farias and Van Roy, 2004. The same approximation was independently proposed by Talluri (2008) under the name of strong affine relaxation and shown to provide tighter upper bounds on the optimal expected revenue than other available methods for the no-choice setting. Also Topaloglu (2009) recently focussed on time- and capacity-dependent bid prices: He proposed a network RM approach based on Lagrangian relaxation, but again without inclusion of choice behavior.
Our key contributions are the following:
- •
We propose a new linear programming approach to approximate dynamic programming that approximates the value function with a nonlinear function of the state vector which is separable over arbitrarily chosen ranges of resource inventory levels. As a special case, we can choose this approximation to be separable over each possible inventory level, which then corresponds to the approximation proposed by Farias and Van Roy (2007), but, in contrast to their approach, our model also accounts for customer choice behavior.
- •
We show that all the linear programs of Liu and van Ryzin, 2008, Zhang and Adelman, 2009, Kunnumkal and Topaloglu, 2008 can be seen as special cases of our linear programming formulation. In particular, for that reason we obtain tighter upper bounds on the objective value than these other approaches and that are asymptotically optimal as time horizon, demand and capacities are linearly scaled up.
- •
We prove that column generation essentially reduces to solving small mixed integer linear programs. Policies for the MNL model with disjoint consideration sets are numerically tested and show significantly improved results.
- •
Due to the larger number of constraints, our approach is considerably more expensive than others if we allow the marginal capacity value estimates to change from any possible inventory level to another. However, we find that sensitivity to inventory levels is most pronounced only relatively close to the departures: Therefore, in order to cut down computational requirements for large networks without much deterioration of the solution quality, we can exploit the flexibility of our model with respect to arbitrary aggregations of inventory levels to solve it with high inventory aggregation at the beginning of the booking horizon, and later to re-solve it with lower aggregation and thus higher accuracy so that we capture the typically more pronounced nonlinearity in inventory levels of the value function closer to the end of the time horizon.
- •
A seemingly new upper bound relationship between the approaches of Zhang and Adelman, 2009, Kunnumkal and Topaloglu, 2008 is shown, namely that the former provides a tighter upper bound on the objective value than the latter.
The paper at hand is organized as follows: In the next section we briefly review the related literature, then in Section 3 we present our model including the required notation followed by the resulting Markov decision process and its equivalent linear programming form in Section 4. We introduce the linear programming models that we compare our approach with in Sections 4.1 Choice-based deterministic LP, 4.2 Alternative deterministic LP, 5 Approximation based on the equivalent LP. Our own approach is derived in Section 5 as well. We show that the column generation subproblem is reducible to a mixed integer linear program in Section 6 and describe bid price policies in Section 7. Finally, we present the computational results in Section 8 and conclude in Section 9.
Section snippets
Literature review
The earliest contributions to single leg RM with choice behavior include (Brumelle et al. (1990) and Belobaba and Weatherford (1996)), amongst others, and for networks the PODS simulation studies by Belobaba and Hopperstad (1999). Zhang and Cooper (2005) consider an inventory control problem of a set of parallel flights including a customer choice model yielding a stochastic optimization problem which is being solved by simulation-based methods. Zhang and Cooper (2009) develop a pricing model
Products
Let our network consist of m resources – that means flight legs in the airline application – and n products. A product consists of a seat on one or several flight legs in combination with a fare class and departure date. Each resource i has a fixed capacity of ci, and the network capacity is given by the corresponding vector c = [c1, … , cm]T. The capacity is homogenous, that means all seats are perfectly substitutable and do not differ, hence allowing us to accommodate all kind of requests from the
Current solution approaches
Let vt(x) denote the expected revenue-to-go from time period t until the final period τ, given the vector x ∈ X of still available resources in the network. The well-known optimality equation for maximizing expected revenue is then given bywith boundary condition vτ+1(x) = 0 for all x. The decision to be made within each time period is which set of products to offer before we can
Approximation based on the equivalent LP
The following linear programming formulation will serve as the starting point of our considerations. It is equivalent to the dynamic program (1) and, for that reason, we denote it by (EQ). The equivalence can be derived from fundamental results of value iteration, see (Powell, 2007), for example.The decision variables are vt(x), for all t, x, and therefore the problem is also intractable for a large state space. The
Solution via column generation
The problem (P) has variables and, for realistic network sizes, cannot be solved in moderate time unless techniques such as column generation are used to deal with problem size. This method builds upon the observation that for large problems most columns never enter the basis matrix and therefore do not need to be stored. Apparently, the main task is then to provide a way of how to find the next column to enter the basis without having to generate the whole coefficient matrix.
Policies
In this section, we address the question of how the solution to (D) can actually be used to obtain a control policy that tells us which set of fares S to offer at any given time t and state x of the network. We use again the notion r(xi) to denote the inventory range of xi on leg i for an arbitrary aggregation.
Numerical results
In this section, we present the results of numerical experiments that shed light on the quality of the upper bounds and performance of policies obtained for our approach, compared with the above mentioned alternative approaches. We consider TISA with different aggregations. The rationale is that we intend to demonstrate the obtainable gains by splitting up the inventory while balancing the computational effort required to solve (P). Our numerical examples provide a framework of what
Conclusion and future research
In the context of quantity-based network revenue management, we presented a linear programming approach to approximate dynamic programming with nonlinear approximation of the value function with the specific feature that it incorporates both customer choice behavior as well as estimates of marginal capacity values that depend on time and resource inventory level. As a result of the improved approximation, we obtain a better estimate of the opportunity cost, which is reflected in provably
References (33)
- et al.
Mathematical programming models for revenue management under customer choice
European Journal of Operational Research
(2010) - et al.
Pricing substitutable flights in airline revenue management
European Journal of Operational Research
(2009) Dynamic bid prices in revenue management
Operations Research
(2007)- Belobaba, P.P., Hopperstad, C., 1999. Boeing/MIT simulation study: PODS results update. In: 1999 AGIFORS Reservations...
- et al.
Comparing decision rules that incorporate customer diversion in perishable asset revenue management situations
Decision Sciences
(1996) - et al.
Neuro-Dynamic Programming
(1996) - et al.
The science of revenue management when passengers purchase the lowest available fare
Journal of Revenue and Pricing Management
(2004) - et al.
Convex Optimization
(2004) - et al.
Allocation of airline seats between stochastically dependent demands
Transportation Science
(1990) - et al.
Cutting Big M down to size
Interfaces
(1990)
An overview of research revenue management: current issues and future research
International Journal of Revenue Management
The linear programming approach to approximate dynamic programming
Operations Research
On constraint sampling in the linear programming approach to approximate dynamic programming
Mathematics of Operations Research
A refined deterministic linear program for the network revenue management problem with customer choice behavior
Naval Research Logistics
Cited by (85)
Reductions of non-separable approximate linear programs for network revenue management
2023, European Journal of Operational ResearchAn Event-Triggered Adaptive Dynamic Programming Method for Large-Scale HVAC Systems
2023, IFAC-PapersOnLineAn interactive booking-limit control for passenger railway revenue management
2021, Journal of Rail Transport Planning and ManagementCitation Excerpt :However, even the network problem can be easily formulated as a dynamic programming model similar to the single-resource problem, its exact optimization is practically impossible because of the well-known “curse of dimensionality” in dynamic programming models. With this concern, many approximate methods for the network problem have been proposed in the past two decades (e.g., de Boer et al., 2002; Liu and van Ryzin, 2008; Bront et al., 2009; Meissner and Strauss, 2012). The most useful outcome of these approximation methods are estimates of opportunity cost (or dual values on resources), which are then used to construct bid prices or booking limits.
A dynamic programming framework for optimal delivery time slot pricing
2021, European Journal of Operational ResearchCitation Excerpt :This DP was initially devised in the fashion industry (Gallego & van Ryzin, 1994), but subsequently adopted and refined by the transportation sector and the attended home delivery industry (Yang et al., 2016). This formulation could be thought of as an instance of a network revenue management problem with customer choice, which finds various applications, e.g. in transportation, hospitality and appointment scheduling problems (see Meissner & Strauss, 2012; Sauré, Patrick, Tyldesley, & Puterman, 2012; Zhang & Adelman, 2009). To find the (approximately) optimal delivery slot prices, we need to compute the value function (at least approximately) for all states and times.
Simulation-based integrated optimization of nesting policy and booking limits for revenue management
2020, Computers and Industrial EngineeringA grey-layered ANP based decision support model for analyzing strategies of resilience in electronic supply chains
2020, Engineering Applications of Artificial Intelligence