Assortment Optimization Under Variants of the Nested Logit Model

Davis, James M.; Gallego, Guillermo; Topaloğlu, Hüseyin

doi:10.1287/opre.2014.1256

Cited by 285 publications

(169 citation statements)

References 30 publications

Supporting

Mentioning

165

Contrasting

Unclassified

Order By: Relevance

“…As a shortcoming, this added complexity results in more complex optimization problems. For example, Davis et al [14] have shown that the optimization of the Nested logit model is in general a NP-hard problem.…”

Section: Related Workmentioning

confidence: 99%

Deep Choice Model Using Pointer Networks for Airline Itinerary Prediction

Mottini¹,

Acuña-Agost²

2017

Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Travel providers such as airlines and on-line travel agents are becoming more and more interested in understanding how passengers choose among alternative itineraries when searching for ights.is knowledge helps them be er display and adapt their o er, taking into account market conditions and customer needs. Some common applications are not only ltering and sorting alternatives, but also changing certain a ributes in real-time (e.g., changing the price). In this paper, we concentrate with the problem of modeling air passenger choices of ight itineraries.is problem has historically been tackled using classical Discrete Choice Modelling techniques. Traditional statistical approaches, in particular the Multinomial Logit model (MNL), is widely used in industrial applications due to its simplicity and general good performance. However, MNL models present several shortcomings and assumptions that might not hold in real applications. To overcome these di culties, we present a new choice model based on Pointer Networks. Given an input sequence, this type of deep neural architecture combines Recurrent Neural Networks with the A ention Mechanism to learn the conditional probability of an output whose values correspond to positions in an input sequence. erefore, given a sequence of di erent alternatives presented to a customer, the model can learn to point to the one most likely to be chosen by the customer. e proposed method was evaluated on a real dataset that combines on-line user search logs and airline ight bookings. Experimental results show that the proposed model outperforms the traditional MNL model on several metrics.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep Choice Model Using Pointer Networks for Airline Itinerary Prediction

Mottini¹,

Acuña-Agost²

2017

Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

show abstract

“…The variants of the assortment optimization problems they study include the introduction of cardinality constraints, the concept of display location based attractiveness, the integration of the pricing decision from a list of possible price levels, the quality consistent pricing and the concept of product precedence constraints. Davis et al (2013b) study the assortment optimization problem, without any constraints on the products, under a nested logit model, where customers first choose a nest of products and then a product from the nest. They identify special cases where the problem is polynomially solvable and they develop tractable methods to obtain the assortments for the NP-hard cases of the problem.…”

Section: Related Literaturementioning

confidence: 99%

The concept and impact analysis of a flexible mobility on demand system

Atasoy

Ikeda

Song

et al. 2015

Transportation Research Part C: Emerging Technologies

134

View full text Add to dashboard Cite

This paper introduces an innovative transportation concept called Flexible Mobility on Demand (FMOD), which provides personalized services to passengers. FMOD is a demand responsive system in which a list of travel options is provided in real-time to each passenger request. The system provides passengers with flexibility to choose from a menu that is optimized in an assortment optimization framework. For operators, there is flexibility in terms of vehicle allocation to different service types: taxi, shared-taxi and mini-bus. The allocation of the available fleet to these three services is carried out dynamically and based on demand and supply so that vehicles can change roles during the day. The FMOD system is built based on a choice model and consumer surplus is taken into account in order to improve the passenger satisfaction. Furthermore, profits of the operators are expected to increase since the system adapts to changing demand patterns. In this paper, we introduce the concept of FMOD and present preliminary simulation results that quantify the added value of this system.

show abstract

“…From the point of view of mathematics, when given a data matrix, to optimize the terrain is network by minimizing the following objective function [32][33][34][35].…”

Section: Mathematical Optimization For Image Processing Tasksmentioning

confidence: 99%