| 1. |
- Gao, Kun, 1993, et al.
(författare)
-
Extrapolation-enhanced model for travel decision making: An ensemble machine learning approach considering behavioral theory
- 2021
-
Ingår i: Knowledge-Based Systems. - : Elsevier BV. - 0950-7051. ; 218
-
Tidskriftsartikel (refereegranskat)abstract
- Modeling individuals’ travel decision making in terms of choosing transport modes, route and departure time for daily activities is an indispensable component for transport system optimization and management. Conventional approaches of modeling travel decision making suffer from presumed model structures and parametric specifications. Emerging machine learning algorithms offer data-driven and non-parametric solutions for modeling travel decision making but encounter extrapolation issues (i.e., disability to predict scenarios beyond training samples) due to neglecting behavioral mechanisms in the framework. This study proposes an extrapolation-enhanced approach for modeling travel decision making, leveraging the complementary merits of ensemble machine learning algorithms (Random Forest in our study) and knowledge-based decision-making theory to enhance both predictive accuracy and model extrapolation. The proposed approach is examined using three datasets about travel decision making, including one estimation dataset (for cross-validation) and two test datasets (for model extrapolation tests). Especially, we use two test datasets containing extrapolated choice scenarios with features that exceed the ranges of training samples, to examine the predictive ability of proposed models in extrapolated choice scenarios, which have hardly been investigated by relevant literature. The results show that both proposed models and the direct application of Random Forest (RF) can give quite good predictive accuracy (around 80%) in the estimation dataset. However, RF has a deficient predictive ability in two test datasets with extrapolated choice scenarios. In contrast, the proposed models provide substantially superior predictive performances in the two test datasets, indicating much stronger extrapolation capacity. The model based on the proposed framework could improve the precision score by 274.93% than the direct application of RF in the first test dataset and by 21.9% in the second test dataset. The results indicate the merits of the proposed approach in terms of prediction power and extrapolation ability as compared to existing methods.
|
|
| 2. |
- Gao, Kun, 1993, et al.
(författare)
-
Quantifying economic benefits from free-floating bike-sharing systems: A trip-level inference approach and city-scale analysis
- 2021
-
Ingår i: Transportation Research Part A: Policy and Practice. - : Elsevier BV. - 0965-8564. ; 144, s. 89-103
-
Tidskriftsartikel (refereegranskat)abstract
- Despite many qualitative discussions about the benefits of free-floating bike-sharing systems (FFBS), high-resolution and quantitative assessments about the economic benefits of FFBS for users are absent. This study proposes an innovative trip-level inference approach for quantifying the economic benefits of FFBS, leveraging massive FFBS transaction data, the emerging multimodal routing Application Programming Interface from online navigators and travel choice modeling. The proposed approach is able to analyze the economic benefit for every single bike-sharing trip and investigate the spatiotemporal heterogeneity in the economic benefits from FFBS. An empirical analysis in Shanghai is conducted using the proposed approach. The estimated saved travel time, cost, and economic benefit due to using FFBS per trip are estimated to be 9.95 min, 3.64 CNY, and 8.68 CNY-eq, respectively. The annual saved travel time, cost, and economic benefits from FFBS in Shanghai are estimated to be 17.665 billion min, 6.463 billion CNY, and 15.410 billion CNY-eq, respectively. The relationships between economic benefits from FFBS and built environment factors in different urban contexts are quantitatively examined using Multiple Linear Regression to explain the spatial heterogeneity in the economic benefits of FFBS. The outcomes provide a useful tool for evaluating the benefits of shared mobility systems, insights into the users’ economic benefit from using FFBS from per-trip, aggregated and spatial perspective, as well as its influencing factors. The results could efficiently support the scientific planning, operation and policy making concerning FFBS in different urban contexts.
|
|
| 3. |
- Gao, Kun, 1993, et al.
(författare)
-
Spatial heterogeneity in distance decay of using bike sharing: An empirical large-scale analysis in Shanghai
- 2021
-
Ingår i: Transportation Research Part D: Transport and Environment. - : Elsevier BV. - 1361-9209. ; 94
-
Tidskriftsartikel (refereegranskat)abstract
- Distance decay is a vital aspect for modeling spatial interactions of human movements and an indispensable input for land use planning and travel demand prediction models. Although many studies have investigated the usage demand of bike-sharing systems in an area, research investigating the distance decay patterns of using dockless bike-sharing systems (DLBS) from a spatially heterogeneous perspective based on large-scale datasets is lacking. This study firstly utilizes massive transaction record data from DLBS in Shanghai of China and online map navigator Application Programming Interface to empirically estimate the distance decay patterns of using DLBS and reveal the spatial heterogeneity in distance decay of using DLBS across different urban contexts. Afterward, this study examines the mechanism of spatial heterogeneity in distance decay, leveraging multiple data resources including Point of Interest (POI) data, demographic data, and road network data. The associations among the distance decay of using DLBS with built environment factors are investigated by multiple linear regression. Results indicate that factors such as population density, land use entropy, branch road density, and metro station density are significantly related to larger distance decay of using DLBS, while factors such as commercial land use ratio, industrial land use ratio, and motorway density are significantly linked to smaller distance decay in Shanghai. Lastly, we further employ an adaptative geographically weighted regression to investigate the spatial divergences of the influences of built environment factors on distance decay. Results reveal notably distinct and even inverse influences of a built environment factor on the distance decay of using DLBS in different urban contexts. The findings provide insights into the distance decay patterns of using DLBS in different urban contexts and their interactions with the built environment, which can support accurate planning and management of sustainable DLBS as per specific urban characteristics.
|
|
| 4. |
- Li, Aoyong, 1993, et al.
(författare)
-
High-resolution assessment of environmental benefits of dockless bike-sharing systems based on transaction data
- 2021
-
Ingår i: Journal of Cleaner Production. - : Elsevier BV. - 0959-6526. ; 296
-
Tidskriftsartikel (refereegranskat)abstract
- Dockless bike-sharing systems (DLBS) have gained much popularity due to their environmentally friendly features. This study puts forward a distinctive framework for assessing the environmental influences of DLBS in high resolution based on DLBS transaction data. The proposed framework firstly estimates the transport mode substituted by DLBS for each recorded bike-sharing trip by utilizing the route planning techniques of online maps and a well-calibrated discrete choice model. Afterward, greenhouse gases (GHG) emission reductions in every recorded DLBS trip are quantified using Life Cycle Analysis. The proposed framework is applied to an empirical dataset from Shanghai, China. The empirical results reveal that the substitution rates of DLBS to different transport modes have substantial spatiotemporal variances and depend strongly on travel contexts, highlighting the necessity of analyzing the environmental impacts of DLBS at the trip level. Moreover, each DLBS trip is estimated to save an average 80.77 g CO2-eq GHG emissions versus than the situations without DLBS in Shanghai. The annual reduced GHG emissions from DLBS are estimated to be 117 kt CO2-eq, which is substantial and equals to the yearly GHG emissions of over 25,000 typical gasoline passenger vehicles. Additionally, the associations among built environments and GHG emission reductions from DLBS are quantitatively investigated to shed light on the spatial variances in the environmental impacts of DLBS. The results can efficiently support the benefit-cost analysis, planning, and management of DLBS.
|
|
| 5. |
- Gao, Kun, 1993, et al.
(författare)
-
Cumulative prospect theory coupled with multi-attribute decision making for modeling travel behavior
- 2021
-
Ingår i: Transportation Research Part A: Policy and Practice. - : Elsevier BV. - 0965-8564. ; 148, s. 1-21
-
Tidskriftsartikel (refereegranskat)abstract
- This study proposes an approach for modeling travel behavior under uncertainty coupling Cumulative Prospect Theory (CPT) with Multi-attribute Decision Making (MADM) theory. CPT is utilized to depict travelers’ evaluations of each attribute, and MADM describes the process of making tradeoffs among multiple conflicting criteria. Divergent perception principles for different attributes are considered in the proposed framework. The proposed approach is utilized for an empirical analysis concerning mode shift behavior for commuting in Shanghai of China, based on data collected by stated preference surveys. Results show that the proposed approach outperforms conventional methods in terms of model performances and behavioral revelations. Empirical results demonstrate that sensitivity to gains and losses in cost and travel time are divergent in mode shift behavior. More importantly, it is found that travelers underestimate the occurrence chances of low-probability travel time and overestimate the occurrence changes of high-probability travel time in mode shift behavior, which is contrary to the findings from economics. Travelers show substantial loss aversion features as well. The heterogeneity in the value functions of CPT is investigated to shed light on differences in the evaluation process among individuals. Results reveal quite different empirical CPT parameters and behavioral mechanisms in mode shift behavior as compared to monetary experiments in economics. It highlights the importance of empirical estimations in various travel choice contexts to essentially understand travel behavior mechanisms, rather than arbitrary usage of findings from economics.
|
|
| 6. |
- Gao, Kun, 1993, et al.
(författare)
-
Diverging effects of subjective prospect values of uncertain time and money
- 2021
-
Ingår i: Communications in Transportation Research. - : Elsevier BV. - 2772-4247. ; 1
-
Tidskriftsartikel (refereegranskat)abstract
- Studies from behavioral economics show that the subjective prospect value of money has diminishing sensitivity to losses/gains, represented by an S-shape, and this has been applied in representing the subjective prospect value of time in many transportation studies such as travel behavior modeling and network equilibrium. In this study, we demonstrate that the prospect value of time has an increasing sensitivity to losses/gains and can be represented by an Ϩ-shape, which contrasts that of money. We further explain the rationality of this surprising finding based on psychological and behavioral theories and discuss extensive practical implications. The correlations between sensitivities to gains and losses in terms of magnitude are revealed as well to shed light on potential underlying correlated behavioral principles. Substantial loss-aversion features are observed in the empirical analysis, supporting endowment effects. Implications of the findings on decision-making and other areas that utilize time as a key indicator have been discussed. The findings may revolutionize many research areas that utilize time as a key indicator such as transportation engineering.
|
|
| 7. |
- Gao, Kun, 1993, et al.
(författare)
-
Examining nonlinear and interaction effects of multiple determinants on airline travel satisfaction
- 2021
-
Ingår i: Transportation Research Part D: Transport and Environment. - : Elsevier BV. - 1361-9209. ; 97
-
Tidskriftsartikel (refereegranskat)abstract
- Improving passengers’ satisfaction is crucial for airline industry and requires in-depth understandings regarding the complex effects of various factors. This study investigates the importance, complex nonlinear effects and interaction effects of various factors (including passenger characteristics and service attributes) on airline travel satisfaction in data-driven manners leveraging machine-learning (ML) approaches. The results show that ML algorithms such as Random Forest have superiority in modeling airline travel satisfaction as compared to conventional logistic regressions. The quantitative importance of various factors is estimated and compared to reveal key determinants of passengers’ satisfaction using permutation-based importance and accumulated local effect analysis. More importantly, results suggest that the main effects of service attributes present piecewise nonlinear patterns. There are piecewise interaction effects between passenger characteristics and service attributes and among service attributes on airline travel satisfaction. Practical implications on efficient and cost-effective measures of promoting satisfaction are derived and discussed based on the findings.
|
|
| 8. |
- Gao, Kun, 1993, et al.
(författare)
-
Modeling Measurements Towards Effect of Past Behavior on Travel Behavior
- 2021
-
Ingår i: Smart Innovation, Systems and Technologies. - Singapore : Springer Singapore. - 2190-3026 .- 2190-3018. ; 231, s. 141-157
-
Konferensbidrag (refereegranskat)abstract
- The inertia effect of past behavior has attracted attention in the travel behavioral literature because of its bearing on travel choice modeling. Several measurements have been proposed to model the inertia effects. However, no consensus concerning appropriate modelling methods is reached, which leads to potential biases in analysis. The study aims to conduct a comprehensive investigation of modeling measurements regarding inertia effects of past behavior from the perspectives of estimation, behavioral indications and predictions. Differing from existing literature that only focused on estimation performance, we examine the performances of different methods in predictions and behavioral interpretations. To our best knowledge, these aspects are not investigated in the literature based on empirical data. The necessary information for constructing the measurements, underlying consumption, significance in estimation, behaviorally implausible issue, performances in estimation and predictions for these measurements are all compared based on behavioral data. The results shed lights on performances and suitability of different measurements for inertia effects in terms of estimation, behavioral interpretation and prediction, which support the further investigations of past behavior on travelers’ choice behavior.
|
|
| 9. |
- Xu, Yueru, et al.
(författare)
-
Modeling Commercial Vehicle Drivers’ Acceptance of Forward Collision Warning System
- 2021
-
Ingår i: Smart Innovation, Systems and Technologies. - Singapore : Springer Singapore. - 2190-3026 .- 2190-3018. ; 231, s. 167-180
-
Konferensbidrag (refereegranskat)abstract
- With the development of computer science, Forward Collision Warning (FCW) systems have been installed in various vehicles in order to improve road safety. Previous studies have been conducted to evaluate the acceptance of FCW systems and explore the contributing factors affecting drivers’ attitudes. However, few research studies have focused on the attitudes of commercial vehicle drivers, though commercial vehicle accidents were proved to be more severe than passenger vehicles. This paper tries to examine the attitudes of commercial vehicle drivers toward FCW systems and identify the contributing factors by using a random forests algorithm. FCW data of 24 commercial vehicles were recorded from November 1st to December 21st, 2018 in Jiangsu province. The acceptance rate (FCW records with response) of commercial vehicle drivers for FCW systems is 69.52%. (Acceptance was measured by identifying drivers who reduced their speed in response to a warning from the FCW system.) The accuracy of random forests model is 0.816 after tuning the parameter. In addition, the most important influence variable in this model is vehicle speed with an importance of 0.37. Duration time and warning hour also have significant influence on driver reaction, with values of 0.20 and 0.17, respectively. The results showed that commercial vehicle drivers’ acceptance of an FCW system decreases with the increase of vehicle speed. The response time for most cases is timely, usually within 2 s. And the response percentage is higher during daytime than at night. These regularities may be attributable to the larger size and heavier weight of commercial vehicles. The results of this study can help researchers to better understand the behavior of commercial vehicle drivers and to develop more effective FCW systems for commercial vehicles.
|
|
| 10. |
- Xue, Qingwen, et al.
(författare)
-
Driving Style Recognition Incorporating Risk Surrogate by Support Vector Machine
- 2021
-
Ingår i: Smart Innovation, Systems and Technologies. - Singapore : Springer Singapore. - 2190-3026 .- 2190-3018. ; 231, s. 123-131
-
Konferensbidrag (refereegranskat)abstract
- Accurate driving style recognition is a crucial component for advanced driver assistance systems and vehicle control systems to reduce potential rear-end collision risk. This study aims to develop a driving style recognition method incorporating matching learning algorithms and vehicle trajectory data. A risk surrogate, Modified Margin to Collision (MMTC), is proposed to evaluate the collision risk level of each driver’s trajectory. Particularly, the traffic level is considered when labelling the driving style, while it has a great impact on driving preference. Afterwards, each driver’s driving style is labelled based on their collision risk level using the K-means algorithm. Driving behavior features, including acceleration, relative speed, and relative distance, are extracted from vehicle trajectory and processed by time-sequence analysis. Finally, Supporting Vector Machine (SVM) is applied to recognize driving style based on the extracted features and labelled data. The performance of Random Forest (RF), K-Nearest Neighbor (KNN), and Multi-Layer Perceptron (MLP) are also compared with SVM. The “leave-one-out” method is used to validate the performance and effectiveness of the proposed model. The results show that SVM over performs others with 91.7% accuracy. This recognition model could effectively recognize the aggressive driving style, which can better support ADAS.
|
|
