Travel Mode Choice Modeling: Predictive Efficacy between Machine Learning Models and Discrete Choice Model

Background: A complex travel behaviour among users is intertwined with many factors. Traditionally, the exploration in travel mode choice modeling has been dominated by the Discrete Choice model, nonetheless, owing to the advancement in computational techniques, machine learning has gained traction in understanding travel behavior. Aim: This study aims at predicting users’ travel model choice by means of machine learning models against a conventional Discrete Choice Model, i.e., Binary Logistic Regression. Objective: To investigate the comparison between machine learning models, namely Neural Network, Random Forest, Decision Tree, and Support Vector Machine against the Discrete Choice Model (Binary Logistic Regression) in the prediction of travel mode choice amongst Kuantan City. Methodology: The dataset was collected in Kuantan City, Malaysia, through the Revealed/Stated Preferences (RP/SP) Survey. The data collected was split into a ratio of 80:20 for training and testing before evaluating them between the aforesaid models. The hyperparameters of the models were set to default. The performance of the models is evaluated based on classification accuracy. Results: It was shown in the present study that the Neural Network Model is able to attain a higher prediction accuracy as compared to Binary Logistic Regression (Discrete Choice Model) in classifying mode choice of Kuantan users either to choose public transport or private vehicles as daily transportation. Feature importance technique is crucial for identifying the significant features in modelling travel mode choice. It is demonstrated that the Neural Network Model can yield exceptional classification of mode choice up to 73.4% and 72.4% of training and testing data, respectively, by considering the features identified via the feature importance technique, suggesting the viability of the proposed technique in supporting an informed decision. Conclusion: The findings highlight the strengths and limitations of the Machine Learning Technique as well as the Discrete Choice Model in modeling travel mode choice. It was shown that Machine Learning models have the capability to provide better prediction that could assist the urban transportation planning among policymakers. Meanwhile, it could be also demonstrated that the Discrete Choice Model (Binary Logistic Regression) is helpful in getting a better understanding in expressing the inference relationship between variables for improvising the future transportation system.

Classification of Imbalanced Travel Mode Choice to Work Data Using Adjustable SVM Model

The investigation of travel mode choice is an essential task in transport planning and policymaking for predicting travel demands. Typically, mode choice datasets are imbalanced and learning from such datasets is challenging. This study deals with imbalanced mode choice data by developing an algorithm (SVMAK) based on a support vector machine model and the theory of adjusting kernel scaling. The kernel function’s choice was evaluated by applying the likelihood-ratio chi-square and weighting measures. The empirical assessment was performed on the 2017 National Household Travel Survey–California dataset. The performance of the SVMAK model was compared with several other models, including neural networks, XGBoost, Bayesian Network, standard support vector machine model, and some SVM-based models that were previously developed to handle the imbalanced datasets. The SVMAK model outperformed these models, and in some cases improved the accuracy of the minority class classification. For the majority class, the accuracy improvement was substantial. This algorithm can be applied to other tasks in the transport planning domain that deal with uneven data distribution.

A Quarter Acre Pavlova Paradise Lost? The Role of Preferences and Planning in Achieving Urban Sustainability in Wellington, New Zealand

<p>This thesis investigates the scope for compact development to accommodate population growth in Wellington, New Zealand. The topic is particularly significant for New Zealand as the great majority of the population lives in urban areas, historical development has been dominated by low density urban form, and transport and urban form are two of the main domains in which the country can reduce its carbon emissions. The influence of urban planning and residents’ preferences on achieving sustainable outcomes is investigated. Historical and current planning rules and transport policies in the City are analysed to determine their influence on the provision of compact development. Wellington’s transport policy shows a pattern of path dependency: historical decisions to favour car oriented investment have driven subsequent transport investments and influenced the ease of using different transport modes. Planning policies show a similar pattern of path dependency: planning rules enacted in the 1960s endure in present planning despite being packaged with different justifications and regulatory regime. Current planning rules severely restrict infill development in most existing neighbourhoods, which reduces the availability of housing in accessible medium density neighbourhoods and likely increases the cost of this type of housing. A stated choice survey was conducted of 454 residents of Wellington City to investigate the extent to which there is an unmet demand for compact development and alternatives to car travel. The survey held presentation mode constant across two completion modes (internet and door to door with tablet completion), allowing the impacts of recruitment and completion mode to be examined. Survey recruitment mode appeared to influence both response rates and the representativeness of the survey, while completion mode appeared to have little or no impact on survey responses. Using the stated choice survey results, a latent class model was developed to examine the preferences of residents and the trade-offs they are willing to make when choosing where to live. This type of model allows for the identification of preference groups as a means of understanding the diversity of preferences across the population. The study found that there is an unmet demand for medium density, accessible housing, but that affordability is a barrier for households to choose this type of housing. There was also an unmet demand for walking and cycling, with more residents currently driving than would prefer to use this mode, and more residents preferring to walk and cycle to work than currently use these modes. The ability to use a desired travel mode appears to be related to the neighbourhood in which a person lives, with residents of medium and high density neighbourhoods being more likely to use their preferred travel mode. This study also modelled future development trajectories for Wellington based on demand for housing, neighbourhood and transport attributes. This preference based growth model was contrasted with the City’s plan for development over the next 30 years. Comparing the two scenarios, the planning based trajectory performed better than the demand based scenario in terms of both carbon emissions and achieving compact development.</p>

A Quarter Acre Pavlova Paradise Lost? The Role of Preferences and Planning in Achieving Urban Sustainability in Wellington, New Zealand

<p>This thesis investigates the scope for compact development to accommodate population growth in Wellington, New Zealand. The topic is particularly significant for New Zealand as the great majority of the population lives in urban areas, historical development has been dominated by low density urban form, and transport and urban form are two of the main domains in which the country can reduce its carbon emissions. The influence of urban planning and residents’ preferences on achieving sustainable outcomes is investigated. Historical and current planning rules and transport policies in the City are analysed to determine their influence on the provision of compact development. Wellington’s transport policy shows a pattern of path dependency: historical decisions to favour car oriented investment have driven subsequent transport investments and influenced the ease of using different transport modes. Planning policies show a similar pattern of path dependency: planning rules enacted in the 1960s endure in present planning despite being packaged with different justifications and regulatory regime. Current planning rules severely restrict infill development in most existing neighbourhoods, which reduces the availability of housing in accessible medium density neighbourhoods and likely increases the cost of this type of housing. A stated choice survey was conducted of 454 residents of Wellington City to investigate the extent to which there is an unmet demand for compact development and alternatives to car travel. The survey held presentation mode constant across two completion modes (internet and door to door with tablet completion), allowing the impacts of recruitment and completion mode to be examined. Survey recruitment mode appeared to influence both response rates and the representativeness of the survey, while completion mode appeared to have little or no impact on survey responses. Using the stated choice survey results, a latent class model was developed to examine the preferences of residents and the trade-offs they are willing to make when choosing where to live. This type of model allows for the identification of preference groups as a means of understanding the diversity of preferences across the population. The study found that there is an unmet demand for medium density, accessible housing, but that affordability is a barrier for households to choose this type of housing. There was also an unmet demand for walking and cycling, with more residents currently driving than would prefer to use this mode, and more residents preferring to walk and cycle to work than currently use these modes. The ability to use a desired travel mode appears to be related to the neighbourhood in which a person lives, with residents of medium and high density neighbourhoods being more likely to use their preferred travel mode. This study also modelled future development trajectories for Wellington based on demand for housing, neighbourhood and transport attributes. This preference based growth model was contrasted with the City’s plan for development over the next 30 years. Comparing the two scenarios, the planning based trajectory performed better than the demand based scenario in terms of both carbon emissions and achieving compact development.</p>

School Travel Inequity between Students from Public and Private Schools in the City of Shenzhen, China

Understanding school travel inequities and promoting active travel policies more effectively is an increasingly important issue in the international transport policy agenda. Using the dataset of the 2014 Shenzhen primary and secondary school travel survey, this study empirically revealed the permanent residence permit (hukou) system in the context of China shapes the evident inequities between students from public schools and private schools. Students without a legitimated hukou to local areas suffer from more constraints, longer distances, and more time to access private schools which are excluded from the public sponsorship and have disadvantages in geographical locations. Applying the ordered logistic model, this study specifically investigated the influential factors of school commuting travel mode. Household vehicle ownership and travel features (i.e., chauffeuring and home-school distance) have a much more significant role in school travel mode decisions, which largely surpassed the role individual demographic attributes and the school surrounding built environment play. The implications of this study shed light on making more specific strategies for private schools to mitigate mobility inequity imposed on disadvantaged students.