Dynamic Programming Approaches for Estimating and Applying Large-scale Discrete Choice Models

People go through their life making all kinds of decisions, and some of these decisions affect their demand for transportation, for example, their choices of where to live and where to work, how and when to travel and which route to take. Transport related choices are typically time dependent and characterized by large number of alternatives that can be spatially correlated. This thesis deals with models that can be used to analyze and predict discrete choices in large-scale networks. The proposed models and methods are highly relevant for, but not limited to, transport applications. We model decisions as sequences of choices within the dynamic discrete choice framework, also known as parametric Markov decision processes. Such models are known to be difficult to estimate and to apply to make predictions because dynamic programming problems need to be solved in order to compute choice probabilities. In this thesis we show that it is possible to explore the network structure and the flexibility of dynamic programming so that the dynamic discrete choice modeling approach is not only useful to model time dependent choices, but also makes it easier to model large-scale static choices. The thesis consists of seven articles containing a number of models and methods for estimating, applying and testing large-scale discrete choice models. In the following we group the contributions under three themes: route choice modeling, large-scale multivariate extreme value (MEV) model estimation and nonlinear optimization algorithms. Five articles are related to route choice modeling. We propose different dynamic discrete choice models that allow paths to be correlated based on the MEV and mixed logit models. The resulting route choice models become expensive to estimate and we deal with this challenge by proposing innovative methods that allow to reduce the estimation cost. For example, we propose a decomposition method that not only opens up for possibility of mixing, but also speeds up the estimation for simple logit models, which has implications also for traffic simulation. Moreover, we compare the utility maximization and regret minimization decision rules, and we propose a misspecification test for logit-based route choice models. The second theme is related to the estimation of static discrete choice models with large choice sets. We establish that a class of MEV models can be reformulated as dynamic discrete choice models on the networks of correlation structures. These dynamic models can then be estimated quickly using dynamic programming techniques and an efficient nonlinear optimization algorithm. Finally, the third theme focuses on structured quasi-Newton techniques for estimating discrete choice models by maximum likelihood. We examine and adapt switching methods that can be easily integrated into usual optimization algorithms (line search and trust region) to accelerate the estimation process. The proposed dynamic discrete choice models and estimation methods can be used in various discrete choice applications. In the area of big data analytics, models that can deal with large choice sets and sequential choices are important. Our research can therefore be of interest in various demand analysis applications (predictive analytics) or can be integrated with optimization models (prescriptive analytics). Furthermore, our studies indicate the potential of dynamic programming techniques in this context, even for static models, which opens up a variety of future research directions.

On payoff heterogeneity in games with strategic complementarities

Payoff heterogeneity weakens positive feedback in binary choice models intwo ways. First, heterogeneity drives individuals to corners where theyare unaffected by strategic complementarities. Second, aggregate behaviouris smoother than individual behaviour when individuals are heterogeneous.However, this smoothing does not necessarily eliminate positive feedbackor guarantee a unique equilibrium. In games with an unbounded, continuouschoice space, heterogeneity may either weaken or strengthen positive feedback,depending on a simple convexity/concavity condition. We conclude that positivefeedback phenomena derived in representative agent models will often be robustto heterogeneity.

Um estudo sobre a vitimização para a cidade de São Paulo

A study about the victimization in the city of São Paulo. This paper applies the crime economics theory to Brazilian data. Following Becker (1968), Hinderlang et al. (1978) and Cohen et al. (1981), we tested the microeconomic factors that influence crime and victimization. For this end, the two waves of research of victimization of the Instituto Futuro Brasil, 2003 and 2008, were used in an effort to identify the determinants of victimization and police notification, using probit model. The main results suggest the factors which impact significantly the probability of victimization are the demographic characteristics, economic conditions and personal habits. The models of "life style" and "opportunity" seem to have good performance.

“A long-term mortality analysis of subsidized firms in rural areas: an empirical study in the Portuguese Alentejo region”

Studies have demonstrated that public policies to support private firms’ investment have the ability to promote entrepreneurship, but the sustainability of subsidized firms has not often been analysed. This paper aims to examine this dimension specifically through evaluating the mortality of subsidized firms in the long-term. The analysis focuses on a case study of the LEADER+ Programme in the Alentejo region of Portugal. With this purpose, the paper examines the activity status (active or not active) of 154 private, rural, for-profit firms in Alentejo that had received a subsidy to support investment between 2002 and 2008 under the LEADER+ Programme. The methodology is based on binary choice models in order to study the probability of these firms still being active. The explanatory variables used are the following: (1) the characteristics of entrepreneurs and managers’ strategic decisions, (2) firm profile and characteristics, (3) regional economic environment. Data assessment showed that the cumulative mortality rate of firms on 31st December 2013 is over 20 %. Interpretation of the regression model revealed that he probability of firms’ survival increases with higher investment, firm age and regional business concentration, whereas the number of applications made by firms has a negative impact on their survival. So it seems that for subsidized firms the amount of investment is as important as its frequency.

Valuing water resources in developing countries: A semiparametric approach to valuation models

Valuation of projects for the preservation of water resources provides important information to policy makers and funding institutions. Standard contingent valuation models rely on distributional assumptions to provide welfare measures. Deviations from assumed and actual distribution of benefits are important when designing policies in developing countries, where inequality is a concern. This article applies semiparametric methods to obtain estimates of the benefit from a project for the preservation of an important Brazilian river basin. These estimates lead to significant differences from those obtained using the standard parametric approach.

On Graphically Checking Goodness-of-fit of Binary Logistic Regression Models

OBJECTIVES: This paper is concerned with checking goodness-of-fit of binary logistic regression models. For the practitioners of data analysis, the broad classes of procedures for checking goodness-of-fit available in the literature are described. The challenges of model checking in the context of binary logistic regression are reviewed. As a viable solution, a simple graphical procedure for checking goodness-of-fit is proposed. METHODS: The graphical procedure proposed relies on pieces of information available from any logistic analysis; the focus is on combining and presenting these in an informative way. RESULTS: The information gained using this approach is presented with three examples. In the discussion, the proposed method is put into context and compared with other graphical procedures for checking goodness-of-fit of binary logistic models available in the literature. CONCLUSION: A simple graphical method can significantly improve the understanding of any logistic regression analysis and help to prevent faulty conclusions.

Improving the analysis of road pricing acceptability surveys by using hybrid models

Several international studies have analyzed the acceptability of road pricing schemes by means of an attitude survey in combination with the results of a stated choice experiment using both a descriptive analysis and a discrete-choice model with binary choice (?accept? or ?not accept? the toll). However, the use of hybrid discrete choice models constitutes an innovative alternative for integrating subjective attitudes and perceptions deriving from the survey of attitudes with the more objective variables from the stated choice experiment. This paper analyzes the results of applying these models to measure the acceptability of interurban road pricing among different groups of stakeholders (road freight and passenger operators, highway concessionaires, and associations of private car users) with qualitatively significant opinions on road pricing measures. Our results show that hybrid models are better suited to explaining the acceptability of a road pricing scheme by different groups of stakeholders than a separate analysis of the survey of attitudes and a discrete-choice model applied on a stated choice experiment. A particular finding was that the strong psycho-social latent variable of the perception of fairness explains the rejection or acceptance of a toll scheme by road stakeholders.

THE IMPORTANCE OF DISAGREEING: CONTRARIANS AND EXTREMISM IN THE CODA MODEL

In this paper, we study the effects of introducing contrarians in a model of Opinion Dynamics where the agents have internal continuous opinions, but exchange information only about a binary choice that is a function of their continuous opinion, the CODA model. We observe that the hung election scenario that arises when contrarians are introduced in discrete opinion models still happens. However, it is weaker and it should not be expected in every election. Finally, we also show that the introduction of contrarians make the tendency towards extremism of the original model weaker, indicating that the existence of agents that prefer to disagree might be an important aspect and help society to diminish extremist opinions.

An agent-based approach to modelling driver route choice behaviour under the influence of real-time information

This paper presents an agent-based approach to modelling individual driver behaviour under the influence of real-time traffic information. The driver behaviour models developed in this study are based on a behavioural survey of drivers which was conducted on a congested commuting corridor in Brisbane, Australia. Commuters' responses to travel information were analysed and a number of discrete choice models were developed to determine the factors influencing drivers' behaviour and their propensity to change route and adjust travel patterns. Based on the results obtained from the behavioural survey, the agent behaviour parameters which define driver characteristics, knowledge and preferences were identified and their values determined. A case study implementing a simple agent-based route choice decision model within a microscopic traffic simulation tool is also presented. Driver-vehicle units (DVUs) were modelled as autonomous software components that can each be assigned a set of goals to achieve and a database of knowledge comprising certain beliefs, intentions and preferences concerning the driving task. Each DVU provided route choice decision-making capabilities, based on perception of its environment, that were similar to the described intentions of the driver it represented. The case study clearly demonstrated the feasibility of the approach and the potential to develop more complex driver behavioural dynamics based on the belief-desire-intention agent architecture. (C) 2002 Elsevier Science Ltd. All rights reserved.

Binary games with many players

We examine a problem with n players each facing the same binary choice. One choice is superior to the other. The simple assumption of competition - that an individual's payoff falls with a rise in the number of players making the same choice, guarantees the existence of a unique symmetric equilibrium (involving mixed strategies). As n increases, there are two opposing effects. First, events in the middle of the distribution - where a player finds itself having made the same choice as many others - become more likely, but the payoffs in these events fall. In opposition, events in the tails of the distribution - where a player finds itself having made the same choice as few others - become less likely, but the payoffs in these events remain high. We provide a sufficient condition (strong competition) under which an increase in the number of players leads to a reduction in the equilibrium probability that the superior choice is made.

Strategy-proof social choice

This paper surveys the literature on strategy-proofness from a historical perspective. While I discuss the connections with other works on incentives in mechanism design, the main emphasis is on social choice models. This article has been prepared for the Handbook of Social Choice and Welfare, Volume 2, Edited by K. Arrow, A. Sen and K. Suzumura

Choice by sequential procedures

We propose a rule of decision-making, the sequential procedure guided byroutes, and show that three influential boundedly rational choice models can be equivalentlyunderstood as special cases of this rule. In addition, the sequential procedure guidedby routes is instrumental in showing that the three models are intimately related. We showthat choice with a status-quo bias is a refinement of rationalizability by game trees, which, inturn, is also a refinement of sequential rationalizability. Thus, we provide a sharp taxonomyof these choice models, and show that they all can be understood as choice by sequentialprocedures.

A dynamic choice model of hybrid behavior in the attribute-space

This paper presents a dynamic choice model in the attributespace considering rational consumers that discount the future. In lightof the evidence of several state-dependence patterns, the model isfurther extended by considering a utility function that allows for thedifferent types of behavior described in the literature: pure inertia,pure variety seeking and hybrid. The model presents a stationaryconsumption pattern that can be inertial, where the consumer only buysone product, or a variety-seeking one, where the consumer buys severalproducts simultane-ously. Under the inverted-U marginal utilityassumption, the consumer behaves inertial among the existing brands forseveral periods, and eventually, once the stationary levels areapproached, the consumer turns to a variety-seeking behavior. An empiricalanalysis is run using a scanner database for fabric softener andsignificant evidence of hybrid behavior for most attributes is found,which supports the functional form considered in the theory.

An enhanced concave program relaxation for choice network revenue management

The network choice revenue management problem models customers as choosing from an offer-set, andthe firm decides the best subset to offer at any given moment to maximize expected revenue. The resultingdynamic program for the firm is intractable and approximated by a deterministic linear programcalled the CDLP which has an exponential number of columns. However, under the choice-set paradigmwhen the segment consideration sets overlap, the CDLP is difficult to solve. Column generation has beenproposed but finding an entering column has been shown to be NP-hard. In this paper, starting with aconcave program formulation based on segment-level consideration sets called SDCP, we add a class ofconstraints called product constraints, that project onto subsets of intersections. In addition we proposea natural direct tightening of the SDCP called ?SDCP, and compare the performance of both methodson the benchmark data sets in the literature. Both the product constraints and the ?SDCP method arevery simple and easy to implement and are applicable to the case of overlapping segment considerationsets. In our computational testing on the benchmark data sets in the literature, SDCP with productconstraints achieves the CDLP value at a fraction of the CPU time taken by column generation and webelieve is a very promising approach for quickly approximating CDLP when segment consideration setsoverlap and the consideration sets themselves are relatively small.

Framework for Skew-Probit Links in Binary Regression

We review several asymmetrical links for binary regression models and present a unified approach for two skew-probit links proposed in the literature. Moreover, under skew-probit link, conditions for the existence of the ML estimators and the posterior distribution under improper priors are established. The framework proposed here considers two sets of latent variables which are helpful to implement the Bayesian MCMC approach. A simulation study to criteria for models comparison is conducted and two applications are made. Using different Bayesian criteria we show that, for these data sets, the skew-probit links are better than alternative links proposed in the literature.