Matching Items (7)
Filtering by
- All Subjects: Transportation
- All Subjects: spatial statistics
- All Subjects: Spatial linear mixed model
- Creators: Khoeini, Sara
- Status: Published
Description
As urban populations become increasingly dense, massive amounts of new 'big' data that characterize human activity are being made available and may be characterized as having a large volume of observations, being produced in real-time or near real-time, and including a diverse variety of information. In particular, spatial interaction (SI) data - a collection of human interactions across a set of origins and destination locations - present unique challenges for distilling big data into insight. Therefore, this dissertation identifies some of the potential and pitfalls associated with new sources of big SI data. It also evaluates methods for modeling SI to investigate the relationships that drive SI processes in order to focus on human behavior rather than data description.
A critical review of the existing SI modeling paradigms is first presented, which also highlights features of big data that are particular to SI data. Next, a simulation experiment is carried out to evaluate three different statistical modeling frameworks for SI data that are supported by different underlying conceptual frameworks. Then, two approaches are taken to identify the potential and pitfalls associated with two newer sources of data from New York City - bike-share cycling trips and taxi trips. The first approach builds a model of commuting behavior using a traditional census data set and then compares the results for the same model when it is applied to these newer data sources. The second approach examines how the increased temporal resolution of big SI data may be incorporated into SI models.
Several important results are obtained through this research. First, it is demonstrated that different SI models account for different types of spatial effects and that the Competing Destination framework seems to be the most robust for capturing spatial structure effects. Second, newer sources of big SI data are shown to be very useful for complimenting traditional sources of data, though they are not sufficient substitutions. Finally, it is demonstrated that the increased temporal resolution of new data sources may usher in a new era of SI modeling that allows us to better understand the dynamics of human behavior.
A critical review of the existing SI modeling paradigms is first presented, which also highlights features of big data that are particular to SI data. Next, a simulation experiment is carried out to evaluate three different statistical modeling frameworks for SI data that are supported by different underlying conceptual frameworks. Then, two approaches are taken to identify the potential and pitfalls associated with two newer sources of data from New York City - bike-share cycling trips and taxi trips. The first approach builds a model of commuting behavior using a traditional census data set and then compares the results for the same model when it is applied to these newer data sources. The second approach examines how the increased temporal resolution of big SI data may be incorporated into SI models.
Several important results are obtained through this research. First, it is demonstrated that different SI models account for different types of spatial effects and that the Competing Destination framework seems to be the most robust for capturing spatial structure effects. Second, newer sources of big SI data are shown to be very useful for complimenting traditional sources of data, though they are not sufficient substitutions. Finally, it is demonstrated that the increased temporal resolution of new data sources may usher in a new era of SI modeling that allows us to better understand the dynamics of human behavior.
ContributorsOshan, Taylor Matthew (Author) / Fotheringham, A. S. (Thesis advisor) / Farmer, Carson J.Q. (Committee member) / Rey, Sergio S.J. (Committee member) / Nelson, Trisalyn (Committee member) / Arizona State University (Publisher)
Created2017
Description
To reduce the environmental burden of transport, previous studies have resorted on solutions that accentuate towards techno-economical pathways. However, there is growing evidence that transport behaviors, lifestyle choices, and the role of individuals' attitudes/perceptions are considered influential factors in shaping households’ engagement with sustainable technologies in the face of environmental crises. The objective of this dissertation is to develop multidimensional econometric model systems to explore complex relationships that can help us understand travel behaviors' implications for transport and household energy use. To this end, the second chapter of this dissertation utilizes the latent segmentation approach to quantify and unravel the relationship between attitudes and behaviors while recognizing the presence of unobserved heterogeneity in the population. It was found that two-thirds of the population fall in the causal structure where behavioral experiences are shaping attitudes, while for one-third attitudes are shaping behaviors. The findings have implications on the energy-behavior modeling paradigm and forecasting household energy use. Building on chapter two, the third chapter develops an integrated modeling framework to explore the factors that influence the adoption of on-demand mobility services and electric vehicle ownership while placing special emphasis on attitudes/perceptions. Results indicated that attitudes and values significantly affect the use of on-demand transportation services and electric vehicle ownership, suggesting that information campaigns and free trials/demonstrations would help advance towards the sustainable transportation future and decarbonize the transport sector. The integrated modeling framework is enhanced, in chapter four, to explore the interrelationship between transport and residential energy consumption. The findings indicated the existence of small but significant net complimentary relationships between transport and residential energy consumption. Additionally, the modeling framework enabled the comparison of energy consumption patterns across market segments. The resulting integrated transport and residential energy consumption model system is utilized, in chapter fifth, to shed light on the overall household energy footprint implications of shifting vehicle/fuel type choices. Results indicated that electric vehicles are driven as much as gasoline vehicles are. Interestingly, while an increase in residential energy consumption was observed with the wide-scale adoption of electric vehicles, the total household energy use decreased, indicating benefits associated with transportation electrification.
ContributorsSharda, Shivam (Author) / Pendyala, Ram M. (Thesis advisor) / Khoeini, Sara (Committee member) / Grimm, Kevin J. (Committee member) / Chester, Mikhail V. (Committee member) / Garikapati, Venu M. (Committee member) / Arizona State University (Publisher)
Created2021
Description
The past two decades have been marked by disruptions in the way transportation is provided to society. Examples are carsharing, ridehailing services, and electric scooters. Understanding how sensitive travel behavior is during transportation disruptions is a key part of planning for the future of transportation. While the effects of people's attitudes and perceptions on travel behavior and choices have been studied in the past, their role in response to disruptions remains under explored. This dissertation explores the effect of attitudes on travel behavior and perceptions for two distinct disruptions: the advent of autonomous vehicles (AVs) and the COVID-19 pandemic. Before diving into such elaborate relationships, it is important to understand how attitudinal data is collected and measured. Thus, a study of the effects of different survey methods on the collection of attitudes towards transportation disruptions is performed. This dissertation finds that having a favorable perception of AVs is the most important factor in defining one’s willingness to use them. More importantly, those who only heard about AVs without knowing much about them were actually less likely to have a favorable perception when compared to those who never heard of AVs prior to the survey, reinforcing the need for thoughtful education and awareness initiatives. Additionally, gender also played an important role in expectations about the AV Future: not only are women less interested in using AVs as a pooled ride service, but also that the effect of attitudes on defining that choice was different for men and women. Regarding the COVID-19 pandemic, two different attitudes towards COVID were identified: concern about the effects of the COVID-19 response, and concern about the health effects of the coronavirus. Both shaped the ways people traveled, and how often they did so. These findings reinforce the need for the broad collection of attitudinal data and the incorporation of such parameters on future travel forecasting.
ContributorsCapasso da Silva, Denise (Author) / Pendyala, Ram M (Thesis advisor) / Khoeini, Sara (Thesis advisor) / Salon, Deborah (Committee member) / Thompson, Marilyn (Committee member) / Arizona State University (Publisher)
Created2021
Description
With rapid advances in technology development and public adoption, it is crucial to understand how these services will shape the future of travel depending on the extent to which people will use these services; impact the transportation and infrastructure systems such as changes in the use of transit and active modes of travel; and influence how technology developers create and update these transportation technologies to better serve people’s mobility needs. This dissertation explores how two major emerging services, namely ridehailing services and autonomous vehicles (AVs), will be used in the future when they are widely available and vastly used, and how they may impact the transportation infrastructure and societal travel patterns. The four proposed chapters use comprehensive quantitative and qualitative methods to explore the status of these technologies from theory, through robust modeling frameworks, to practice, by investigating the recent AV pilot deployments in real-world settings. In the second chapter, it was found that increased frequency of ridehailing use is significantly associated with a decrease in bus usage, suggesting that ridehailing functions more as a substitute for buses than as a complement and implying that transit agencies should explore ways to incorporate ridehailing services in their plans to enhance transit usage. Next, the third chapter showed that interest in using AVs for running errands had a positive and significant effect on AV ownership intent, even after accounting for a host of variables. The fourth chapter depicted how ridehailing experiences have a considerable effect on the willingness to ride AV-based services in both private and shared modes, suggesting that experience is crucial for future adoption of these services. Then, two recent real-world AV experiences are explored in the fifth chapter. Lessons learned from these experiments reinforced the importance of first-hand experiences in promoting AV awareness and trustworthiness, potentially leading to greater degrees of adoption. Finally, the results and discussions presented in this dissertation strengthen the body of literature on key emerging transportation technologies and inform policymakers and stakeholders to properly prepare cities and the public to welcome these technologies into our transportation system in an efficient, equitable, and complementary way.
ContributorsMagassy, Tassio Bezerra (Author) / Pendyala, Ram M (Thesis advisor) / Khoeini, Sara (Committee member) / Polzin, Steven E (Committee member) / Salon, Deborah (Committee member) / Arizona State University (Publisher)
Created2023
Description
Spatial regression is one of the central topics in spatial statistics. Based on the goals, interpretation or prediction, spatial regression models can be classified into two categories, linear mixed regression models and nonlinear regression models. This dissertation explored these models and their real world applications. New methods and models were proposed to overcome the challenges in practice. There are three major parts in the dissertation.
In the first part, nonlinear regression models were embedded into a multistage workflow to predict the spatial abundance of reef fish species in the Gulf of Mexico. There were two challenges, zero-inflated data and out of sample prediction. The methods and models in the workflow could effectively handle the zero-inflated sampling data without strong assumptions. Three strategies were proposed to solve the out of sample prediction problem. The results and discussions showed that the nonlinear prediction had the advantages of high accuracy, low bias and well-performed in multi-resolution.
In the second part, a two-stage spatial regression model was proposed for analyzing soil carbon stock (SOC) data. In the first stage, there was a spatial linear mixed model that captured the linear and stationary effects. In the second stage, a generalized additive model was used to explain the nonlinear and nonstationary effects. The results illustrated that the two-stage model had good interpretability in understanding the effect of covariates, meanwhile, it kept high prediction accuracy which is competitive to the popular machine learning models, like, random forest, xgboost and support vector machine.
A new nonlinear regression model, Gaussian process BART (Bayesian additive regression tree), was proposed in the third part. Combining advantages in both BART and Gaussian process, the model could capture the nonlinear effects of both observed and latent covariates. To develop the model, first, the traditional BART was generalized to accommodate correlated errors. Then, the failure of likelihood based Markov chain Monte Carlo (MCMC) in parameter estimating was discussed. Based on the idea of analysis of variation, back comparing and tuning range, were proposed to tackle this failure. Finally, effectiveness of the new model was examined by experiments on both simulation and real data.
In the first part, nonlinear regression models were embedded into a multistage workflow to predict the spatial abundance of reef fish species in the Gulf of Mexico. There were two challenges, zero-inflated data and out of sample prediction. The methods and models in the workflow could effectively handle the zero-inflated sampling data without strong assumptions. Three strategies were proposed to solve the out of sample prediction problem. The results and discussions showed that the nonlinear prediction had the advantages of high accuracy, low bias and well-performed in multi-resolution.
In the second part, a two-stage spatial regression model was proposed for analyzing soil carbon stock (SOC) data. In the first stage, there was a spatial linear mixed model that captured the linear and stationary effects. In the second stage, a generalized additive model was used to explain the nonlinear and nonstationary effects. The results illustrated that the two-stage model had good interpretability in understanding the effect of covariates, meanwhile, it kept high prediction accuracy which is competitive to the popular machine learning models, like, random forest, xgboost and support vector machine.
A new nonlinear regression model, Gaussian process BART (Bayesian additive regression tree), was proposed in the third part. Combining advantages in both BART and Gaussian process, the model could capture the nonlinear effects of both observed and latent covariates. To develop the model, first, the traditional BART was generalized to accommodate correlated errors. Then, the failure of likelihood based Markov chain Monte Carlo (MCMC) in parameter estimating was discussed. Based on the idea of analysis of variation, back comparing and tuning range, were proposed to tackle this failure. Finally, effectiveness of the new model was examined by experiments on both simulation and real data.
ContributorsLu, Xuetao (Author) / McCulloch, Robert (Thesis advisor) / Hahn, Paul (Committee member) / Lan, Shiwei (Committee member) / Zhou, Shuang (Committee member) / Saul, Steven (Committee member) / Arizona State University (Publisher)
Created2020
Description
Motorcycle fatalities have been increasing at a faster rate than the number of motorcycles being registered in the United States. There is limited analysis on the causes of fatal motorcycle crashes, specifically regarding different demographics, certain driver behavior, and various crash characteristics. It is important to be aware of how these factors relate to each other during a fatal motorcycle crash. This analysis focuses on these factors and explores potential steps to decrease motorcycle fatality rates using research and data from the Fatality Analysis Reporting System (FARS) from the National Highway Traffic Safety Administration (NHTSA), and data from the National Household Travel Survey (NHTS). Based on this data, there are noticeable trends between different genders and age groups. According to the analysis, males have a higher fatality rate than females, and their fatal crashes tend to involve multiple driver infractions such as drinking, speeding, not wearing a helmet, and driving without a license. Similarly, younger drivers have a higher fatality rate than older drivers, and their fatal crashes tend to involve multiple driver infractions. Although older drivers involved in fatal crashes usually drive more cautiously, they tend to be involved in single-vehicle crashes more often than younger drivers. Moving forward, implementing certain training programs directed towards particular demographics has the potential to decrease motorcycle rider fatalities.
ContributorsMoran, Sarah Elizabeth (Co-author) / Santilli, Amy (Co-author) / Pendyala, Ram (Thesis director) / Khoeini, Sara (Committee member) / Industrial, Systems & Operations Engineering Prgm (Contributor, Contributor) / Barrett, The Honors College (Contributor)
Created2019-05
Description
Approximately 1% of the total working population within the United States bikes as their primary mode of commute. Due to recent increased in bicycle facilities as well as a focus on alternative modes of transport, understanding the motivations and type of people who bike to work is important in order to encourage new users.
In this project, a literature review was completed as well as data analysis of the National Household Travel Survey (NHTS) in order to find specific populations to target. Using these target populations, it is suggested that advertising and workplace encouragement occur to persuade more people to bike to work. Through data analysis it was found that the most impactful variables were the region of the country, gender, population density, and commute distance. Bicycle commuters statistically had fewer vehicles in their households and drove less miles annually.
There were five main target groups found through this analysis; people who bike for other reasons besides work and live in a city with more than 4,000 people per square mile, young professionals between 19-39, women in regions with separated bicycle facilities, those with low vehicle availability, and environmentally conscious individuals. Working to target these groups through advertising campaigns to encourage new users, as well as increasing and improving bicycle facilities, will help create more new bicyclists.
In this project, a literature review was completed as well as data analysis of the National Household Travel Survey (NHTS) in order to find specific populations to target. Using these target populations, it is suggested that advertising and workplace encouragement occur to persuade more people to bike to work. Through data analysis it was found that the most impactful variables were the region of the country, gender, population density, and commute distance. Bicycle commuters statistically had fewer vehicles in their households and drove less miles annually.
There were five main target groups found through this analysis; people who bike for other reasons besides work and live in a city with more than 4,000 people per square mile, young professionals between 19-39, women in regions with separated bicycle facilities, those with low vehicle availability, and environmentally conscious individuals. Working to target these groups through advertising campaigns to encourage new users, as well as increasing and improving bicycle facilities, will help create more new bicyclists.
ContributorsImbus, Eileen Elizabeth (Author) / Khoeini, Sara (Thesis director) / Pendyala, Ram (Committee member) / Civil, Environmental and Sustainable Eng Program (Contributor) / Barrett, The Honors College (Contributor)
Created2019-05