The accuracy of the multinomial logit model as an approximation to the multinomial probit model of travel demand

The multinomial probit model of travel demand is considerably more general but much less tractable than the better-known multinomial logit model. In an effort to determine the effects of using the relatively simple logit model in situations where the assumptions of probit modeling are satisfied but those of logit modeling are not, the accuracy of the multinomial logit model as an approximation to a variety of three-alternative probit models has been evaluated. Multinomial logit can give highly erroneous estimates of the choice probabilities of multinomial probit models. However, logit models appear to give asymptotically accurate estimates of the ratios of the coefficients of the systematic components of probit utility functions, even when the logit choice probabilities differ greatly from the probit ones. Large estimation data sets are not necessarily needed to enable likelihood ratio tests to distinguish three-alternative probit models from logit models that give seriously erroneous estimates of the probit choice probabilities. Inclusion of alternative-specific dummy variables in logit utility functions cannot be relied upon to reduce significantly the errors of logit approximations to the choice probabilities of probit models whose utility functions do not contain the dummies.     

Methods of determining choice probability in utility maximising multiple alternative models

Methods of estimating choice probabilities between multiple alternatives are described, within the context of various methods of specifying the degree of correlation between the alternatives. A utility maximising framework is assumed. The models described are based either on the logit family or multinomial probit. Two new methods of analytical approximation for the multinomial probit model are introduced, which appear to show significant advantages over the traditional Clark method. Some comparative results of choice probability estimates are presented to support this contention. The conclusion discusses the relative usefulness of different choice estimation methods for different types of problem.     

Estimating choice probabilities among nested alternatives

In several travel choice situations (e.g. automobile ownership level and trip frequency) the alternatives available to an individual randomly chosen from the population exhibit some internal choice-related ranking: the choice of a given alternative implies that all lower-ranked alternatives have been chosen. Such alternatives are referred to as “nested”. This paper presents a model for estimating choice probabilities among nested alternatives. The model is devised from the well known logit model and uses existing logit maximum-likelihood estimation techniques (and computer packages). The approach is shown to be more attractive than the multinomial logit and linear regression models, from a theoretical point of view, yet cheaper than the multinomial probit model. The model is developed in a disaggregate, utility maximization framework. An example application, estimating probabilities of trip frequencies by elderly individuals is presented.     

Embedding risk attitude and decision weights in non-linear logit to accommodate time variability in the value of expected travel time savings

In recent years we have seen important extensions of logit models in behavioural research such as incorporation of preference and scale heterogeneity, attribute processing heuristics, and estimation of willingness to pay (WTP) in WTP space. With rare exception, however, a non-linear treatment of the parameter set to allow for behavioural reality, such as embedded risk attitude and perceptual conditioning of occurrence probabilities attached to specific attributes, is absent. This is especially relevant to the recent focus in travel behaviour research on identifying the willingness to pay for reduced travel time variability, which is the source of estimates of the value of trip reliability that has been shown to take on an increasingly important role in project appraisal. This paper incorporates, in a generalised non-linear (in parameters) logit model, alternative functional forms for perceptual conditioning (known as probability weighting) and risk attitude in the utility function to account for travel time variability, and then derives an empirical estimate of the willingness to pay for trip time variability-embedded travel time savings as an alternative to separate estimates of time savings and trip time reliability. We illustrate the richness of the approach using a stated choice data set for commuter choice between unlabelled attribute packages. Statistically significant risk attitude parameters and parameters underlying decision weights are estimated for multinomial logit and mixed multinomial logit models, along with values of expected travel time savings.     

The valuation of commuter travel time savings for car drivers: evaluating alternative model specifications

The empirical valuation of travel time savings is a derivative of the ratio of parameter estimates in a discrete choice model. The most common formulation (multinomial logit) imposes strong restrictions on the profile of the unobserved influences on choice as represented by the random component of a preference function. As we progress our ability to relax these restrictions we open up opportunities to benchmark the values derived from simple (albeit relatively restrictive) models. In this paper we contrast the values of travel time savings derived from multinomial logit and alternative specifications of mixed (or random parameter) logit models. The empirical setting is urban car commuting in six locations in New Zealand. The evidence suggests that less restrictive choice model specifications tend to produce higher estimates of values of time savings compared to the multinomial logit model; however the degree of under-estimation of multinomial logit remains quite variable, depending on the context.     

Transition choice probabilities and welfare analysis in random utility models with imperfect before–after correlation

Welfare in random utility models is used to be analysed on the basis of only the expectation of the compensating variation. De Palma and Kilani (De Palma, A., Kilani, K., 2011. Transition choice probabilities and welfare analysis in additive random utility models. Economic Theory 46(3), 427–454) have developed a framework for conditional welfare analysis which provides analytic expressions of transition choice probabilities and associated welfare measures. The contribution is of practical relevance in transportation because it allows to compute shares of shifters and non-shifters and attribute benefits to them in a rigorous way. In De Palma and Kilani (2011) the usual assumption of unchanged random terms before and after is made.The paper generalises the framework for conditional welfare analysis to cases of imperfect before–after association of the random terms. The joint before–after distribution of the random terms is introduced with postulated properties in terms of marginal distributions and covariance matrix. Analytic expressions, based on the probability density function and the cumulative distribution function of the joint before–after distribution, and simulation procedures for computation of the transition choice probabilities and the conditional expectations of the compensating variation are provided. Results are specialised for multinomial logit and probit. In the case without income effects, it is proved that the unconditional expectation of the compensating variation depends only on the marginal distributions.The theory is illustrated by a numerical example which refers to a multinomial logit applied to the choice of the transport mode with two specifications, one without and one with income effects. Results show that transition probabilities and conditional welfare measures are affected significantly by the assumption on the before–after correlation. The variability in the transition probabilities across transitions tends to decrease as the before–after correlation decreases. In the extreme case of independent random terms, the conditional expectations of the compensating variation tend to be close to the unconditional expectation.     

A theoretical and empirical model of trip chaining behavior

This paper addresses the theoretical and empirical issues involved in modeling complex travel patterns. Existing models have the shortcoming of not representing the interdependencies among trip links in trip chains with multiple non-home stops. A theoretical model based on utility theory and explicitly accounting for the trade-offs involved in the choice of multiple-stop chains is developed. Using this theoretical model, utility maximizing conditions for a household's choice of a daily travel pattern are derived. The optimum travel pattern is described in terms of the number of chairs (tours) traveled on a given day and in terms of the number of stops (sojourns) made on each of those chains. For a given household, the form of the optimum pattern is a function of the transportation expenditures (time, cost) required to reach potential destinations. Constraints on the conditions of optimality due to the limited and discrete nature of travel pattern alternatives are also considered. Parameters of the general utility function were estimated empirically using actual travel data derived from a home interview survey taken in Washington, D.C. The multinomial logit model is used to relate utility scores for the alternative travel patterns to choice probabilities. The resulting parameter estimates agree with theoretical expectations and with empirical results obtained in other studies. In order to demonstrate the empirical and theoretical implications of the model, forecasts for various transportation policies (e.g., gasoline price increases, transit fare reductions), as made by this model and by other less complex models, are compared. The results of these comparisons indicate the need for expanding the scope of existing travel forecasting models to explicit considerations of trip chaining behavior.     

Does the choice model method and/or the data matter?

The opportunity to have seven data sets associated with a stated choice experiment that are very similar in content and design is rare, and provides an opportunity to look in detail at the empirical evidence within and between each data set in the context of a range of discrete choice estimation methods, from multinomial logit to latent class to scale multinomial logit to mixed logit, and the most general model, generalized mixed multinomial logit that accounts for preference and scale heterogeneity. Given the problems associated with data from different countries and time periods, we estimate separate models for each data set, obtaining values of travel time savings that are then updated post estimation to a common dollar for comparative purposes. We also pooled all data sets for a scaled MNL model, treating each data set as a set of three separate utility expressions, but linked to the other data sets through scale heterogeneity. This is not behaviourally appropriate with MNL, latent class or mixed logit. The main question investigated is whether there exists greater synergy in the willingness to pay evidence within model form across data sets compared to across model forms within data sets. The evidence suggests that there is a relatively greater convergence of evidence across the choice models, with the exception of generalized mixed logit, after controlling for data set differences; and there is strong evidence to suggest that differences between data sets do matter.     

A nested logit model of automobile holdings for one vehicle households

This paper presents a model of automobile choice by single vehicle households. This effort is distinguished from previous disaggregate automobile holdings models primarily by the use of the nested logit model rather than the more restrictive multinomial logit model. We present a 2-step estimation technique that provides consistent and asymptotically efficient parameter estimates, yet is tractable for very large choice sets. Using disaggregate data on 237 one-vehicle households we estimate the unknown parameters on an automobile choice model containing 785 individual makes, models and vintages of passenger vehicles.     

A class of RUM choice models that includes the model in which the utility has logistic distributed errors

A class of random utility maximization (RUM) models is introduced. For these RUM models the utility errors are the sum of two independent random variables, where one of them follows a Gumbel distribution. For this class of RUM models an integral representation of the choice probability generating function has been derived which is substantially different from the usual integral representation arising from the RUM theory. Four types of models belonging to the class are presented. Thanks to the new integral representation, a closed-form expression for the choice probability generating function for these four models may be easily obtained. The resulting choice probabilities are fairly manageable and this fact makes the proposed models an interesting alternative to the logit model. The proposed models have been applied to two samples of interurban trips in Japan and some of them yield a better fit than the logit model. Finally, the concavity of the log-likelihood of the proposed models with respect to the utility coefficients is also analyzed.     

Reconsidering the multinomial probit model

Recently developed computational methods have greatly reduced the difficulty of estimating multinomial probit models and may soon make multinomial probit a computationally feasible option in applied travel demand modeling. This paper discusses some of the benefits and costs that are associated with the use of multinomial probit in demand modeling. It is argued that although there are situations in which multinomial probit is essential for achieving a satisfactory model, most problems with existing demand models are unlikely to be mitigated by the use of multinomial probit.     

An integrated model for discrete and continuous decisions with application to vehicle ownership,type and usage choices

This paper proposes an integrated econometric framework for discrete and continuous choice dimensions. The model system is applied to the problem of household vehicle ownership, type and usage. A multinomial probit is used to estimate household vehicle ownership, a multinomial logit is used to estimate the vehicle type (class and vintage) choices, and a regression is used to estimate the vehicle usage decisions. Correlation between the discrete (number of vehicles) and the continuous (total annual miles traveled) parts is captured with a full variance–covariance matrix of the unobserved factors. The model system is estimated using Simulated Log-Likelihood methods on data extracted from the 2009 US National Household Travel Survey and a secondary dataset on vehicle characteristics. Model estimates are applied to evaluate changes in vehicle holding and miles driven, in response to the evolution of social societies, living environment and transportation policies.     

Microscopic decision model for pedestrian route choice at signalized crosswalks

In this paper, two‐tier mathematical models were developed to simulate the microscopic pedestrian decision‐making process of route choice at signalized crosswalks. In the first tier, a discrete choice model was proposed to predict the choices of walking direction. In the second tier, an exponential model was calibrated to determine the step size in the chosen direction. First, a utility function was defined in the first‐tier model to describe the change of utility in response to deviation from a pedestrian's target direction and the conflicting effects of neighboring pedestrians. A mixed logit model was adopted to estimate the effects of the explanatory variables on the pedestrians' decisions. Compared with the standard multinomial logit model, it was shown that the mixed logit model could accommodate the heterogeneity. The repeated observations for each pedestrian were grouped as panel data to ensure that the parameters remained constant for individual pedestrians but varied among the pedestrians. The mixed logit model with panel data was found to effectively address inter‐pedestrian heterogeneity and resulted in a better fit than the standard multinomial logit model. Second, an exponential model in the second tier was proposed to further determine the step size of individual pedestrians in the chosen direction; it indicates the change in walking speed in response to the presence of other pedestrians. Finally, validation was conducted on an independent set of observation data in Hong Kong. The pedestrians' routes and destinations were predicted with the two‐tier models. Compared with the tracked trajectories, the average error between the predicted destinations and the observed destinations was within an acceptable margin. Copyright © 2016 John Wiley & Sons, Ltd.     

Unified closed-form expression of logit and weibit and its extension to a transportation network equilibrium assignment

This study proposes a generalized multinomial logit model that allows heteroscedastic variance and flexible utility function shape. The novelty of our approach is that the model is theoretically derived by applying a generalized extreme-value distribution to the random component of utility, while retaining its closed-form expression. In addition, the weibit model, in which the random utility is assumed to follow the Weibull distribution, is a special case of the proposed model. This is achieved by utilizing the q-generalization method developed in Tsallis statistics. Then, our generalized logit model is incorporated into a transportation network equilibrium model. The network equilibrium model with a generalized logit route choice is formulated as an optimization problem for uncongested networks. The objective function includes Tsallis entropy, a type of generalized entropy. The generalization of the Gumbel and Weibull distributions, logit and weibit models, and network equilibrium model are formulated within a unified framework with q-generalization or Tsallis statistics.     

Effects of task complexity and time pressure on activity-travel choices: heteroscedastic logit model and activity-travel simulator experiment

This paper derives, estimates and applies a discrete choice model of activity-travel behaviour that accommodates potential effects of task complexity and time pressure on decision-making. To the best of our knowledge, this is the first time that both factors (task complexity and time pressure) are jointly captured in a discrete choice model. More specifically, our heteroscedastic logit model captures potential impacts of task complexity and time pressure through the scale of the utility of activity-travel options. We collect data using a novel activity-travel simulator experiment that has been specifically designed with the aim of testing our model. Results are in line with expectations, in that higher levels of task complexity and time pressure are found to result in a smaller scale of utility. In other words, higher levels of task complexity and time pressure lead to more random choice behaviour and as a consequence to less pronounced differences in choice probabilities between alternatives. An empirical illustration suggests that choice probability-differences between models that do and those that do not capture these effects, can be very substantial; this in turn suggests that failing to capture the effects of task complexity and time pressure in discrete choice models of activity travel decision-making might lead to serious bias in forecasts of the effects of transport policies.     

Strategic sampling for large choice sets in estimation and application

Many discrete choice contexts in transportation deal with large choice sets, including destination, route, and vehicle choices. Model estimation with large numbers of alternatives remains computationally expensive. In the context of the multinomial logit (MNL) model, limiting the number of alternatives in estimation by simple random sampling (SRS) yields consistent parameter estimates, but estimator efficiency suffers. In the context of more general models, such as the mixed MNL, limiting the number of alternatives via SRS yields biased parameter estimates. In this paper, a new, strategic sampling scheme is introduced, which draws alternatives in proportion to updated choice-probability estimates. Since such probabilities are not known a priori, the first iteration uses SRS among all available alternatives. The sampling scheme is implemented here for a variety of simulated MNL and mixed-MNL data sets, with results suggesting that the new sampling scheme provides substantial efficiency benefits. Thanks to reductions in estimation error, parameter estimates are more accurate, on average. Moreover, in the mixed MNL case, where SRS produces biased estimates (due to violation of the independence of irrelevant alternatives property), the new sampling scheme appears to effectively eliminate such biases. Finally, it appears that only a single iteration of the new strategy (following the initialization step using SRS) is needed to deliver the strategy’s maximum efficiency gains.     

A new random utility model with flexible correlation pattern and closed-form covariance expression: The CoRUM

This paper proposes a new random utility model characterised by a cumulative distribution function (cdf) obtained as a finite mixture of different cdfs. This entails that choice probabilities, covariances and elasticities of this model are also a finite mixture of choice probabilities, covariances and elasticities of the mixing models. As a consequence, by mixing nested logit cdfs, a model is generated with closed-form expressions for choice probabilities, covariances and elasticities and with, potentially, a very flexible correlation pattern. Importantly, the closed-form covariance expression opens up interesting application possibilities in some special choice contexts, like route choice, where prior expectations in terms of the covariance matrix can be formulated.     

The multinomial logit model revisited: A semi-parametric approach in discrete choice analysis

The multinomial logit model in discrete choice analysis is widely used in transport research. It has long been known that the Gumbel distribution forms the basis of the multinomial logit model. Although the Gumbel distribution is a good approximation in some applications such as route choice problems, it is chosen mainly for mathematical convenience. This can be restrictive in many other scenarios in practice. In this paper we show that the assumption of the Gumbel distribution can be substantially relaxed to include a large class of distributions that is stable with respect to the minimum operation. The distributions in the class allow heteroscedastic variances. We then seek a transformation that stabilizes the heteroscedastic variances. We show that this leads to a semi-parametric choice model which links the linear combination of travel-related attributes to the choice probabilities via an unknown sensitivity function. This sensitivity function reflects the degree of travelers’ sensitivity to the changes in the combined travel cost. The estimation of the semi-parametric choice model is also investigated and empirical studies are used to illustrate the developed method.     

A new logit‐artificial neural network ensemble for mode choice modeling: a case study for border transport

Logit model is one of the statistical techniques commonly used for mode choice modeling, while artificial neural network (ANN) is a very popular type of artificial intelligence technique used for mode choice modeling. Ensemble learning has evolved to be very effective approach to enhance the performance for many applications through integration of different models. In spite of this advantage, the use of ANN‐based ensembles in mode choice modeling is under explored. The focus of this study is to investigate the use of aforementioned techniques for different number of transportation modes and predictor variables. This study proposes a logit‐ANN ensemble for mode choice modeling and investigates its efficiency in different situations. Travel between Khobar‐Dammam metropolitan area of Saudi Arabia and Kingdom of Bahrain is selected for mode choice modeling. The travel on this route can be performed mainly by air travel or private vehicle through King Fahd causeway. The results show that the proposed ensemble gives consistently better accuracies than single models for multinomial choice problems irrespective of number of input variables. Copyright © 2015 John Wiley & Sons, Ltd.     

Generation of synthetic datasets for discrete choice analysis

Despite the widespread use of synthetic data in discrete choice analysis, little is known about how the methodology used to generate synthetic datasets influences the properties of parameter estimates and the validity of results based on these estimates. That is, there are two potential sources of biases when using synthetic discrete choice data: (1) bias due to the method used to generate the dataset; and, (2) bias due to parameter estimation. The primary objective of this study is to examine bias due to the underlying data generation method. This study compares three methods for generating synthetic datasets and uses design of experiments and analysis of variance methods to investigate the ability to recover estimates for “true” logsum parameters for nested logit models. The method that uses nested logit probabilities to generate the chosen alternative results in unbiased parameter estimates. The method that is based on Gumbel error component approximations reveals that while the error components themselves are unbiased, subtle empirical identification problems can arise when these error components are combined with synthetically generated utility functions. The method that is based on normal error component approximations reveals that all logsum coefficients are biased upwards; the bias dramatically increases for those nests that have a low choice frequency and is most pronounced for those nests with high correlations among alternatives. Based on the results of the analysis, several recommendations for the generation of synthetic datasets for discrete choice analyses are provided.