Attitudes on Autonomous Vehicle Adoption using Interpretable Gradient Boosting Machine

2019 ◽  
Vol 2673 (11) ◽  
pp. 865-878 ◽  
Author(s):  
Dongwoo Lee ◽  
John Mulrow ◽  
Chana Joanne Haboucha ◽  
Sybil Derrible ◽  
Yoram Shiftan
Keyword(s):  
Panel Data ◽  
Autonomous Vehicles ◽  
Choice Model ◽  
Environmental Concern ◽  
Stated Preference ◽  
Autonomous Vehicle ◽  
Predictive Performance ◽  
Gradient Boosting ◽  
Purchase Cost ◽  
Gradient Boosting Machine

This article applies machine learning (ML) to develop a choice model on three choice alternatives related to autonomous vehicles (AV): regular vehicle (REG), private AV (PAV), and shared AV (SAV). The learned model is used to examine users’ preferences and behaviors on AV uptake by car commuters. Specifically, this study applies gradient boosting machine (GBM) to stated preference (SP) survey data (i.e., panel data). GBM notably possesses more interpretable features than other ML methods as well as high predictive performance for panel data. The prediction performance of GBM is evaluated by conducting a 5-fold cross-validation and shows around 80% accuracy. To interpret users’ behaviors, variable importance (VI) and partial dependence (PD) were measured. The results of VI indicate that trip cost, purchase cost, and subscription cost are the most influential variables in selecting an alternative. Moreover, the attitudinal variables Pro-AV Sentiment and Environmental Concern are also shown to be significant. The article also examines the sensitivity of choice by using the PD of the log-odds on selected important factors. The results inform both the modeling of transportation technology uptake and the configuration and interpretation of GBM that can be applied for policy analysis.

scholarly journals User Automotive Powertrain-Type Choice Model and Analysis Using Neural Networks

2021 ◽  
Vol 13 (2) ◽  
pp. 585
Author(s):  
Fabio Luis Marques dos Santos ◽  
Paolo Tecchio ◽  
Fulvio Ardente ◽  
Ferenc Pekár
Keyword(s):  
Internal Combustion Engine ◽  
User Behavior ◽  
Choice Model ◽  
Combustion Engine ◽  
Stated Preference ◽  
Operating Cost ◽  
Internal Combustion ◽  
Ann Model ◽  
Purchase Cost ◽  
Artificial Neural Network Ann

This paper presents an artificial neural network (ANN) model that simulates user’s choice of electric or internal combustion engine automotive vehicles based on basic vehicle attributes (purchase price, range, operating cost, taxes due to emissions, time to refuel/recharge and vehicle price depreciation), with the objective of analyzing user behavior and creating a model that can be used to support policymaking. The ANN was trained using stated preference data from a survey carried out in six European countries, taking into account petrol, diesel and battery electric automotive vehicle attributes. Model results show that the electric vehicle parameters (especially purchase cost, range and recharge times), as well as the purchase cost of internal combustion engine vehicles, have the most influence on consumers’ vehicle choices. A graphical interface was created for the model, to make it easier to understand the interactions between different attributes and their impacts on consumer choices and thus help policy decisions.

scholarly journals Public Preferences and Willingness to Pay for Shared Autonomous Vehicles Services in Nagoya, Japan

Smart Cities ◽  
2019 ◽  
Vol 2 (2) ◽  
pp. 230-244 ◽  
Author(s):  
Mingyang Hao ◽  
Yanyan Li ◽  
Toshiyuki Yamamoto
Keyword(s):  
Willingness To Pay ◽  
Autonomous Vehicles ◽  
Stated Preference ◽  
Autonomous Vehicle ◽  
Transportation Mode ◽  
Public Preferences ◽  
Socioeconomic Characteristics ◽  
On Demand ◽  
Vehicle Systems ◽  
The Relationship

Shared autonomous vehicle systems are anticipated to offer cleaner, safer, and cheaper mobility services when autonomous vehicles are finally implemented on the roads. The evaluation of people’s intentions regarding shared autonomous vehicle services appears to be critical prior to the promotion of this emerging mobility on demand approach. Based on a stated preference survey in Nagoya, Japan, the preference for shared autonomous vehicle services as well as willingness to pay for these services were examined among 1036 respondents in order to understand the relationship between people’s socioeconomic characteristics and their preferred shared autonomous vehicle services. For this purpose, k-modes clustering technique was selected and six clusters were obtained. Six groups with respect to different interests on shared autonomous vehicle services were clustered. The result of correlation analysis and discussion of willingness to pay on services provided insightful results for the future shared autonomous vehicle services. This study not only aids in revealing the demands of customer different clusters, but also states the prospective needs of users for stakeholders from research, policymaker and industry field, who are preparing to work on promoting shared autonomous vehicle systems, and subsequently, develops an optimum transportation mode by considering both demand and services as a whole.

scholarly journals Comparison of Random Forest and Gradient Boosting Machine Models for Predicting Demolition Waste Based on Small Datasets and Categorical Variables

2021 ◽  
Vol 18 (16) ◽  
pp. 8530
Author(s):  
Gi-Wook Cha ◽  
Hyeun-Jun Moon ◽  
Young-Chan Kim
Keyword(s):  
Random Forest ◽  
Predictive Models ◽  
Model Development ◽  
Predictive Performance ◽  
Construction And Demolition Waste ◽  
Categorical Variables ◽  
Gradient Boosting ◽  
Demolition Waste ◽  
Gradient Boosting Machine ◽  
Generation Information

Construction and demolition waste (DW) generation information has been recognized as a tool for providing useful information for waste management. Recently, numerous researchers have actively utilized artificial intelligence technology to establish accurate waste generation information. This study investigated the development of machine learning predictive models that can achieve predictive performance on small datasets composed of categorical variables. To this end, the random forest (RF) and gradient boosting machine (GBM) algorithms were adopted. To develop the models, 690 building datasets were established using data preprocessing and standardization. Hyperparameter tuning was performed to develop the RF and GBM models. The model performances were evaluated using the leave-one-out cross-validation technique. The study demonstrated that, for small datasets comprising mainly categorical variables, the bagging technique (RF) predictions were more stable and accurate than those of the boosting technique (GBM). However, GBM models demonstrated excellent predictive performance in some DW predictive models. Furthermore, the RF and GBM predictive models demonstrated significantly differing performance across different types of DW. Certain RF and GBM models demonstrated relatively low predictive performance. However, the remaining predictive models all demonstrated excellent predictive performance at R2 values > 0.6, and R values > 0.8. Such differences are mainly because of the characteristics of features applied to model development; we expect the application of additional features to improve the performance of the predictive models. The 11 DW predictive models developed in this study will be useful for establishing detailed DW management strategies.

scholarly journals Shared Autonomous Vehicles Effect on Vehicle-Km Traveled and Average Trip Duration

2018 ◽  
Vol 2018 ◽  
pp. 1-10 ◽  
Author(s):  
Ana T. Moreno ◽  
Andrzej Michalski ◽  
Carlos Llorca ◽  
Rolf Moeckel
Keyword(s):  
Metropolitan Area ◽  
Autonomous Vehicles ◽  
Online Survey ◽  
Travel Demand ◽  
Stated Preference ◽  
Autonomous Vehicle ◽  
Trip Duration ◽  
Conventional Vehicle ◽  
Willingness To Use ◽  
The Impact

Intermediate modes of transport, such as shared vehicles or ride sharing, are starting to increase their market share at the expense of traditional modes of car, public transport, and taxi. In the advent of autonomous vehicles, single occupancy shared vehicles are expected to substitute at least in part private conventional vehicle trips. The objective of this paper is to estimate the impact of shared autonomous vehicles on average trip duration and vehicle-km traveled in a large metropolitan area. A stated preference online survey was designed to gather data on the willingness to use shared autonomous vehicles. Then, commute trips and home-based other trips were generated microscopically for a synthetic population in the greater Munich metropolitan area. Individuals who traveled by auto were selected to switch from a conventional vehicle to a shared autonomous vehicle subject to their willingness to use them. The effect of shared autonomous vehicles on urban mobility was assessed through traffic simulations in MATSim with a varying autonomous taxi fleet size. The results indicated that the total traveled distance increased by up to 8% after autonomous fleets were introduced. Current travel demand can still be satisfied with an acceptable waiting time when 10 conventional vehicles are replaced with 4 shared autonomous vehicles.

scholarly journals Who Is Willing to Share Their AV? Insights about Gender Differences among Seven Countries

2021 ◽  
Vol 13 (9) ◽  
pp. 4769
Author(s):  
Amalia Polydoropoulou ◽  
Ioannis Tsouros ◽  
Nikolas Thomopoulos ◽  
Cristina Pronello ◽  
Arnór Elvarsson ◽  
...  
Keyword(s):  
Willingness To Pay ◽  
Autonomous Vehicles ◽  
Stated Preference ◽  
Autonomous Vehicle ◽  
Mixed Logit ◽  
Travel Behaviour ◽  
User Preferences ◽  
Traffic Conditions ◽  
Study Results ◽  
Willingness To Adopt

The introduction of shared autonomous vehicles into the transport system is suggested to bring significant impacts on traffic conditions, road safety and emissions, as well as overall reshaping travel behaviour. Compared with a private autonomous vehicle, a shared automated vehicle (SAV) is associated with different willingness-to-adopt and willingness-to-pay characteristics. An important aspect of future SAV adoption is the presence of other passengers in the SAV—often people unknown to the cotravellers. This study presents a cross-country exploration of user preferences and WTP calculations regarding mode choice between a private non-autonomous vehicle, and private and shared autonomous vehicles. To explore user preferences, the study launched a survey in seven European countries, including a stated-preference experiment of user choices. To model and quantify the effect of travel mode attributes and socio-demographic characteristics, the study employs a mixed logit model. The model results were the basis for calculating willingness-to-pay values for all countries and travel modes, and provide insight into the significant heterogeneous, gender-wise effect of cotravellers in the choice to use an SAV. The study results highlight the importance of analysis of the effect of SAV attributes and shared-ride conditions on the future acceptance and adoption rates of such services.

scholarly journals Forecasting the Walking Assistance Rehabilitation Level of Stroke Patients Using Artificial Intelligence

Diagnostics ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1096
Author(s):  
Kanghyeon Seo ◽  
Bokjin Chung ◽  
Hamsa Priya Panchaseelan ◽  
Taewoo Kim ◽  
Hyejung Park ◽  
...  
Keyword(s):  
Predictive Performance ◽  
Medical Rehabilitation ◽  
Gradient Boosting ◽  
Support Vector ◽  
Prescription Data ◽  
Automated Classification ◽  
Classification Models ◽  
Light Gradient ◽  
Gradient Boosting Machine ◽  
Tree Models

Cerebrovascular accidents (CVA) cause a range of impairments in coordination, such as a spectrum of walking impairments ranging from mild gait imbalance to complete loss of mobility. Patients with CVA need personalized approaches tailored to their degree of walking impairment for effective rehabilitation. This paper aims to evaluate the validity of using various machine learning (ML) and deep learning (DL) classification models (support vector machine, Decision Tree, Perceptron, Light Gradient Boosting Machine, AutoGluon, SuperTML, and TabNet) for automated classification of walking assistant devices for CVA patients. We reviewed a total of 383 CVA patients’ (1623 observations) prescription data for eight different walking assistant devices from five hospitals. Among the classification models, the advanced tree-based classification models (LightGBM and tree models in AutoGluon) achieved classification results of over 90% accuracy, recall, precision, and F1-score. In particular, AutoGluon not only presented the highest predictive performance (almost 92% in accuracy, recall, precision, and F1-score, and 86.8% in balanced accuracy) but also demonstrated that the classification performances of the tree-based models were higher than that of the other models on its leaderboard. Therefore, we believe that tree-based classification models have potential as practical diagnosis tools for medical rehabilitation.

scholarly journals A Data-Driven Approach to Characterize the Impact of Connected and Autonomous Vehicles on Traffic Flow

2020 ◽  
Author(s):  
Amir Bahador Parsa ◽  
Ramin Shabanpour ◽  
Abolfazl Mohammadian ◽  
Joshua Auld ◽  
Thomas Stephens
Keyword(s):  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Gradient Boosting ◽  
Analysis Tool ◽  
Urban Structure ◽  
K Nearest Neighbors ◽  
Network Characteristics ◽  
Extreme Gradient Boosting ◽  
Data Driven Approach ◽  
The Impact

The current study aims to present a model to characterize changes in network traffic flows as a result of implementing connected and autonomous vehicle (CAV) technology based on traffic network and built-environment characteristics. To develop such a model, first, POLARIS agent-based modeling platform is used to predict changes in average daily traffic (ADT) under CAVs scenario in the road network of Chicago metropolitan area as the dependent variable of the model. Second, a comprehensive set of variables and indicators representing network characteristics and urban structure patterns are generated. Three machine learning models namely K-Nearest neighbors, Random Forest, and eXtreme Gradient Boosting are developed and validated to establish the relationship between network characteristics and changes in ADT under CAVs scenario. The estimated models are found to yield acceptable performance. In addition, SHapley Additive exPlanations (SHAP) analysis tool is employed to investigate the impact of important features on changes in ADT, which discloses the most important link properties, network features, and demographic information in predicting change in ADT under the analyzed CAVs scenario.

Vision-Based Navigation of Autonomous Vehicles in Roadway Environments with Unexpected Hazards

2019 ◽  
Vol 2673 (12) ◽  
pp. 494-507 ◽  
Author(s):  
Mhafuzul Islam ◽  
Mashrur Chowdhury ◽  
Hongda Li ◽  
Hongxin Hu
Keyword(s):  
Object Detection ◽  
Autonomous Vehicles ◽  
Autonomous Vehicle ◽  
Semantic Segmentation ◽  
Steering Wheel ◽  
Potential Hazard ◽  
Driving System ◽  
Hazardous Object ◽  
Vision Based Navigation ◽  
Navigational System

Vision-based navigation of autonomous vehicles primarily depends on the deep neural network (DNN) based systems in which the controller obtains input from sensors/detectors, such as cameras, and produces a vehicle control output, such as a steering wheel angle to navigate the vehicle safely in a roadway traffic environment. Typically, these DNN-based systems in the autonomous vehicle are trained through supervised learning; however, recent studies show that a trained DNN-based system can be compromised by perturbation or adverse inputs. Similarly, this perturbation can be introduced into the DNN-based systems of autonomous vehicles by unexpected roadway hazards, such as debris or roadblocks. In this study, we first introduce a hazardous roadway environment that can compromise the DNN-based navigational system of an autonomous vehicle, and produce an incorrect steering wheel angle, which could cause crashes resulting in fatality or injury. Then, we develop a DNN-based autonomous vehicle driving system using object detection and semantic segmentation to mitigate the adverse effect of this type of hazard, which helps the autonomous vehicle to navigate safely around such hazards. We find that our developed DNN-based autonomous vehicle driving system, including hazardous object detection and semantic segmentation, improves the navigational ability of an autonomous vehicle to avoid a potential hazard by 21% compared with the traditional DNN-based autonomous vehicle driving system.

Sign in / Sign up

Export Citation Format

Share Document