The benefits of allowing heteroscedastic stochastic distributions in multiple discrete-continuous choice models

This paper investigates the benefits of incorporating heteroscedastic stochastic distributions in random utility maximization-based multiple discrete-continuous (MDC) choice models. To this end, first, a Multiple Discrete-Continuous Heteroscedastic Extreme Value (MDCHEV) model is formulated to allow heteroscedastic extreme value stochastic distributions in MDC models. Next, an empirical analysis of individuals' daily time use choices is carried out using data from the National Household Travel Survey (NHTS) for three geographical regions in Florida. A variety of different MDC model structures are estimated: (a) the Multiple Discrete-Continuous Extreme Value (MDCEV) model with independent and identically distributed (IID) extreme value error structure, (b) the MDCHEV model, (c) the mixed-MDCEV model that allows heteroscedasticity by mixing a heteroscedastic distribution over an IID extreme value kernel, (d) the MDC generalized extreme value (MDCGEV) model that allows inter-alternative correlations using the multivariate extreme value error structure, (e) the mixed-MDCEV model that allows inter-alternative correlations using common mixing distributions across choice alternatives, and (f) the mixed-MDCEV model that allows both heteroscedasticity and inter-alternative correlations. Among all these model structures, the MDCHEV model provided the best fit to the current empirical data. Further, heteroscedasticity was prominent while no significant inter-alternative correlations were found. Specifically, the MDCHEV parameter estimates revealed the significant presence of heteroscedasticity in the random utility components of different activity type choice alternatives. On the other hand, the MDCEV model resulted in inferior model fit and systematic discrepancies between the observed and predicted distributions of time allocations, which can be traced to the thick right tail of the type-1 extreme value distribution. The MDCHEV model addressed these issues to a considerable extent by allowing tighter stochastic distributions for certain choice alternatives, thanks to its accommodation of heteroscedasticity among random utility components. Furthermore, spatial transferability assessments using different transferability metrics also suggest that the MDCHEV model clearly outperformed the MDCEV model.

[1]  Chandra R. Bhat,et al.  A multiple discrete–continuous extreme value model: formulation and application to discretionary time-use decisions , 2005 .

[2]  V. Sivaraman,et al.  A Discrete-Continuous Choice Model with Perfect and Imperfect Substitutes , 2015 .

[3]  Greg M. Allenby,et al.  Multiple-Constraint Choice Models with Corner and Interior Solutions , 2011, Mark. Sci..

[4]  Frank S. Koppelman,et al.  TRANSFERABILITY ANALYSIS OF DISAGGREGATE CHOICE MODELS , 1982 .

[5]  Chandra R. Bhat,et al.  A Multiple Discrete-Continuous Nested Extreme Value (MDCNEV) Model: Formulation and Application to Non-worker Activity Time-Use and Timing Behavior on Weekdays , 2010 .

[6]  Xin Ye,et al.  A Hybrid Land Conversion Model Incorporating Multiple End Uses , 2011, Agricultural and Resource Economics Review.

[7]  Peter E. Rossi,et al.  Modeling Consumer Demand for Variety , 2002 .

[8]  C. Bhat.  A heteroscedastic extreme value model of intercity mode choice , 1995 .

[9]  Abdul Rawoof Pinjari,et al.  Analysis of long-distance vacation travel demand in the United States: a multiple discrete–continuous choice framework , 2011 .

[10]  Yeonbae Kim,et al.  A forecast of household ownership and use of alternative fuel vehicles: A multiple discrete-continuous choice approach , 2008 .

[11]  Daniel J. Phaneuf,et al.  Estimation and Welfare Calculations in a Generalized Corner Solution Model with an Application to Recreation Demand , 2000, Review of Economics and Statistics.

[12]  Roger H. von Haefen.  Empirical Strategies for Incorporating Weak Complementarity into Continuous Semand System Models , 2005 .

[13]  T. Wales,et al.  Estimation of consumer demand systems with binding non-negativity constraints☆ , 1983 .

[14]  William H. Press,et al.  Numerical recipes , 1990 .

[15]  Sujan Sikder,et al.  Spatial Transferability of Person-Level Daily Activity Generation and Time Use Models , 2013 .

[16]  Chandra R. Bhat,et al.  A New Estimation Approach for the Multiple Discrete-Continuous Probit (MDCP) Choice Model , 2013 .

[17]  Chandra R. Bhat,et al.  A Self Instructing Course in Mode Choice Modeling: Multinomial and Nested Logit Models , 2006 .

[18]  Chandra R. Bhat,et al.  Allowing for non-additively separable and flexible utility forms in multiple discrete-continuous models , 2013 .

[19]  Eric J. Miller,et al.  Multidimensional Decision Modeling Framework for Built Space Supply , 2013 .

[20]  Khandker Nurul Habib,et al.  Modeling Daily Activity Program Generation Considering Within-Day and Day-to-Day Dynamics in Activity-Travel Behavior , 2007 .

[21]  V. Smith,et al.  Recreation Demand Models , 2005 .

[22]  Functional Forms in Discrete/Continuous Choice Models with General Corner Solutions , 2009 .

[23]  Khandker Nurul Habib,et al.  Modelling daily activity program generation considering within-day and day-to-day dynamics in activity-travel behaviour , 2008 .

[24]  D. Hensher,et al.  Residential self-selection effects in an activity time-use behavior model , 2009 .

[25]  C. Bhat.  A heteroscedastic extreme value model of intercity travel mode choice , 1995 .

[26]  Abdul Rawoof Pinjari,et al.  Generalized extreme value (GEV)-based error structures for multiple discrete-continuous choice models , 2011 .

[27]  Junyi Zhang,et al.  Representing in-home and out-of-home energy consumption behavior in Beijing , 2011 .

[28]  Mark D. Uncles,et al.  Discrete Choice Analysis: Theory and Application to Travel Demand , 1987 .

[29]  Andrew Daly,et al.  Estimating choice models containing attraction variables , 1982 .

[30]  K. Axhausen,et al.  An analysis of weekly out-of-home discretionary activity participation and time-use behavior , 2009 .

[31]  R. Pendyala,et al.  Multiple Discrete-Continuous Model of Activity Type Choice and Time Allocation for Home-Based Nonwork Tours , 2013 .

[32]  G. Parsons,et al.  Estimation and Welfare Analysis With Large Demand Systems , 2004 .

[33]  Chandra R. Bhat,et al.  The maximum approximate composite marginal likelihood (MACML) estimation of multinomial probit-based unordered response choice models , 2011 .

[34]  Chandra R. Bhat,et al.  The Impact of Demographics, Built Environment Attributes, Vehicle Characteristics, and Gasoline Prices on Household Vehicle Holdings and Use , 2009 .

[35]  W. Michael Hanemann,et al.  Functional Forms in Discrete/Continuous Choice Models With General Corner Solution , 2008 .

[36]  Kay W. Axhausen,et al.  Exploring Variation Properties of Time Use Behavior Based on a Multilevel Multiple Discrete-Continuous Extreme Value Model , 2009 .

[37]  Chandra R. Bhat,et al.  Accommodating Multiple Constraints in the Multiple-Discrete Continuous Extreme Value (MDCEV) Choice Model , 2012 .

[38]  F. Vito.  Residential Energy Demand: a Multiple Discrete-Continuous Extreme Value Model using Italian Expenditure Data , 2012 .

[39]  Daniel J. Phaneuf,et al.  Kuhn-Tucker Demand System Approaches to Non-Market Valuation , 2005 .

[40]  Chandra R. Bhat,et al.  Workshop report: recent advances on modeling multiple discrete-continuous choices , 2013 .

[41]  C. Bhat,et al.  Household Vehicle Type Holdings and Usage: An Application of the Multiple Discrete-Continuous Extreme Value (MDCEV) Model , 2006 .

[42]  C. Bhat.  The multiple discrete-continuous extreme value (MDCEV) model : Role of utility function parameters, identification considerations, and model extensions , 2008 .

[43]  K. Axhausen,et al.  Exploring Variation Properties of Time Use Behavior on the Basis of a Multilevel Multiple Discrete–Continuous Extreme Value Model , 2010 .

[44]  K. Axhausen,et al.  Stated response and multiple discrete-continuous choice models: Analyses of residuals , 2011 .