Discrete Choice Models and Artificial Intelligence Techniques for Predicting the Determinants of Transport Mode Choice--A Systematic Review

Forecasting travel demand requires a grasp of individual decision-making behavior.However,transport mode choice(TMC)is determined by personal and contextual factors that vary from person to person.Numerous characteristics have a substantial impact on travel behavior(TB),which makes it important to take into account while studying transport options.Traditional statistical techniques frequently presume linear correlations,but real-world data rarely follows these presumptions,which may make it harder to grasp the complex interactions.Thorough systematic review was conducted to examine how machine learning(ML)approaches might successfully capture nonlinear correlations that conventional methods may ignore to overcome such challenges.An in-depth analysis of discrete choice models(DCM)and several ML algorithms,datasets,model validation strategies,and tuning techniques employed in previous research is carried out in the present study.Besides,the current review also summarizes DCM and ML models to predict TMC and recognize the determinants of TB in an urban area for different transport modes.The two primary goals of our study are to establish the present conceptual frameworks for the factors influencing the TMC for daily activities and to pinpoint methodological issues and limitations in previous research.With a total of 39 studies,our findings shed important light on the significance of considering factors that influence the TMC.The adjusted kernel algorithms and hyperparameter-optimized ML algorithms outperform the typical ML algorithms.RF(random forest),SVM(support vector machine),ANN(artificial neural network),and interpretable ML algorithms are the most widely used ML algorithms for the prediction of TMC where RF achieved an R2 of 0.95 and SVM achieved an accuracy of 93.18%;however,the adjusted kernel enhanced the accuracy of SVM 99.81%which shows that the interpretable algorithms outperformed the typical algorithms.The sensitivity analysis indicates that the most significant parameters influencing TMC are the age,total trip time,and the number of drivers.

Discrete Choice Analysis of Temporal Factors on Social Network Growth

Social networks like Facebook, X (Twitter), and LinkedIn provide an interaction and communication environment for users to generate and share content, allowing for the observation of social behaviours in the digital world. These networks can be viewed as a collection of nodes and edges, where users and their interactions are represented as nodes and the connections between them as edges. Understanding the factors that contribute to the formation of these edges is important for studying network structure and processes. This knowledge can be applied to various areas such as identifying communities, recommending friends, and targeting online advertisements. Several factors, including node popularity and friends-of-friends relationships, influence edge formation and network growth. This research focuses on the temporal activity of nodes and its impact on edge formation. Specifically, the study examines how the minimum age of friends-of-friends edges and the average age of all edges connected to potential target nodes influence the formation of network edges. Discrete choice analysis is used to analyse the combined effect of these temporal factors and other well-known attributes like node degree (i.e., the number of connections a node has) and network distance between nodes. The findings reveal that temporal properties have a similar impact as network proximity in predicting the creation of links. By incorporating temporal features into the models, the accuracy of link prediction can be further improved.

An empirical study on travel demand management modeling based on discrete choice method

In view of the problem that the requirements of travel demand management and traffic policy-sensitivity are ignored during the establishment process of the travel demand forecasting model, a discrete-choice-based travel demand forecasting model is proposed to demonstrate its applicability to travel demand management. A car-bus discrete choice model is established, including three variables, i. e,, individual socioeconomic characteristics, time, and cost, and the traffic policy-sensitivity is evaluated through two kinds of traffic policies： parking charges and bus priorities. The empirical results show that travel choice is insensitive to the policy of parking charges as 88. 41% of the travelers are insensitive to parking charges; travel choice is, however, sensitive to the policy of bus priorities as 67.70% of the car travelers and 77.02% of the bus travelers are sensitive to bus priorities. The discrete-choice-based travel demand forecasting model is quite policy-sensitive and also has a good adaptability for travel demand management when meeting the basic functions of the demand forecasting model.

Application of discrete choice model in trip mode structure forecast:a case study of Bengbu

In order to find the main factors that influence the urban traffic structure,a relational model between the travelers＇ characteristics and the trip mode choice is built.The data of urban residents＇ characteristics are obtained from statistical data,while the trip mode split data is collected through a trip survey in Bengbu.In addition,the discrete choice model is adopted to build the functional relationship between the mode choice and the travelers＇ personal characteristics,as well as family characteristics and trip characteristics.The model shows that the relationship between the mode split and the personal,as well as family and trip characteristics is stable and changes little as the time changes.Deduced by the discrete model,the mode split result is relatively accurate and can be feasibly used for trip mode structure forecasts.Furthermore,the proposed model can also contribute to find the key influencing factors on trip mode choice,and restructure or optimize the urban trip mode structure.

Comparative study on mode split discrete choice models

Discrete choice model acts as one of the most important tools for studies involving mode split in the context of transport demand forecast. As different types of discrete choice models display their merits and restrictions diversely, how to properly select the specific type among discrete choice models for realistic application still remains to be a tough problem. In this article, five typical discrete choice models for transport mode split are, respectively, discussed, which includes multinomial logit model, nested logit model （NL）, heteroscedastic extreme value model, multinominal probit model and mixed multinomial logit model （MMNL）. The theoretical basis and application attributes of these five models are especially analysed with great attention, and they are also applied to a realistic intercity case of mode split forecast, which results indi- cating that NL model does well in accommodating similarity and heterogeneity across alternatives, while MMNL model serves as the most effective method for mode choice prediction since it shows the highest reliability with the least significant prediction errors and even outperforms the other four models in solving the heterogeneity and similarity problems. This study indicates that conclusions derived from a single discrete choice model are not reliable, and it is better to choose the proper model based on its characteristics.

The effect of online reviews on addressing endogeneity in discrete choice models

This paper investigates the effectiveness of online reviews on addressing price endogeneity issue in an application to consumer demand for smartphone.We consider review variables as the substitutes of unobserved product quality in terms of a scalar variable as seen in previous methods.An aspect-based sentiment classification technique is designed to construct feature-related review variables from millions of review contents.We discuss the performance of review variables both in a hedonic pricing model and a conditional logit discrete choice model.Our results demonstrate that review variables show a good performance either as instruments for price or as explicit control variables in demand models.In detail,the pricing prediction accuracy increases 3.4%,which is considered as a significant improvement in the practice of forecasting.In the discrete choice model,the estimated price coefficient is biased in the positive direction without endogeneity correction.It is adjusted in the expected way after including review variables.The findings indicate that online reviews provide alternative sources of information in dealing with endogeneity in discrete choice models.We also analyze the differences in the preferences and needs of individual consumers to provide some practical implications of marketing.

Estimation of charging demand for electric vehicles by discrete choice models and numerical simulations: Application to a case study in Turin

The electrification of vehicles is considered one of the most important strategies for addressing the issues related to energy dependence and climate change.To meet user needs,electric vehicle(EV)management for charging operations is essential.This study uses modelling and simulation of EV user behaviour to forecast possible scenarios for electric charging in cities and to identify potential management problems and opportunities for improvement of EVs and EV charging infrastructures.The conurbation of Turin was selected as a case study to reproduce realistic scenarios by applying discrete choice modelling based on socio-economic and transport system data.One of objectives of the study was to describe user charging behaviour from a geographic perspective to model where users prefer to charge in the area studied according to the variables that may affect decisions.Another objective was to estimate the number of electric vehicles in Turin and the characteristics of their users,both of which are helpful in understanding electric mobility within a city.Analysing these behavioural issues in a modelling framework can provide a set of tools to compare and evaluate a variety of possible modifications,indicating an adequate network of charging infrastructure to facilitate the diffusion of electric vehicles.

Clicking through the Clickstream: A Novel Statistical Modeling Approach to Improve Information Usage of Clickstream Data by E-Commerce Entities

Success or failure of an E-commerce platform is often reduced to its ability to maximize the conversion rate of its visitors. This is commonly regarded as the capacity to induce a purchase from a visitor. Visitors possess individual characteristics, histories, and objectives which complicate the choice of what platform features that maximize the conversion rate. Modern web technology has made clickstream data accessible allowing a complete record of a visitor’s actions on a website to be analyzed. What remains poorly constrained is what parts of the clickstream data are meaningful information and what parts are accidental for the problem of platform design. In this research, clickstream data from an online retailer was examined to demonstrate how statistical modeling can improve clickstream information usage. A conceptual model was developed that conjectured relationships between visitor and platform variables, visitors’ platform exit rate, boune rate, and decision to purchase. Several hypotheses on the nature of the clickstream relationships were posited and tested with the models. A discrete choice logit model showed that the content of a website, the history of website use, and the exit rate of pages visited had marginal effects on derived utility for the visitor. Exit rate and bounce rate were modeled as beta distributed random variables. It was found that exit rate and its variability for pages visited were associated with site content, site quality, prior visitor history on the site, and technological preferences of the visitor. Bounce rate was also found to be influenced by the same factors but was in a direction opposite to the registered hypotheses. Most findings supported that clickstream data is amenable to statistical modeling with interpretable and comprehensible models.

电子围栏背景下共享单车停车行为影响特征分析

单车规范停放问题影响着共享单车系统的可持续发展,而设置电子围栏是促进单车停放规范化的重要手段。用户在电子围栏背景下规范停车行为的影响机理是制定停放管理政策的基础。基于用户停车选择行为问卷调查的结果,在二项Logit模型的基础上,结合调查数据特征建立固定效应选择模型、非对称选择模型等多个停车行为选择模型,探究各类特征对用户停车行为的影响。结果显示,用户的年龄、职业、收入等社会经济特征,用户在平常使用共享单车的时段、时长、频率及停放困难频率等共享单车出行特征,具体在停车场景下合规停放的绕行距离和违规停放的罚款金额等场景特征,均会显著影响用户合规停放的概率。模型结果可为电子围栏背景下的精细化管理提供用户行为定量分析基础。

旅游纪念品购买偏好及其异质性来源——基于离散选择实验

购买旅游纪念品是旅游体验的重要组成部分,也是旅游目的地商家的关键业务,了解旅游纪念品自身属性对购买决策的影响程度非常重要。文章设计离散选择实验模拟购物情境,收集旅游者选择数据,利用随机参数Logit模型估计,深入分析旅游者对旅游纪念品属性的选择偏好及偏好异质性来源。研究发现:(1)相较于功能属性,旅游者购买旅游纪念品时更看重价值属性,其中,地方性是旅游者最看重的属性,其次是审美性、文化性、礼品性、实用性、便携性、价格;(2)旅游者在审美性和价格的偏好上存在异质性,随着年龄增大、收入增加,审美性、价格对旅游者购买意愿的影响减小,相比男性旅游者,女性旅游者对旅游纪念品的价格更加敏感。

考虑随机用户均衡的区县级电动汽车快充站规划

电动汽车渗透率的提高和城市化进程的发展使得区县级充电设施的建设具有重要意义。综合考虑用户充电需求时空分布和用户充电决策中的有限理性,提出一种面向区县的电动汽车快充站优化规划模型。首先,针对区县用户的出行特征,构建包含外部连接的交通网络拓扑扩展形式,采用出行链仿真捕捉用户快充需求;其次,结合影响用户充电选择的多种因素,建立用户充电决策模型并给出随机用户均衡条件;然后,建立以投资运行成本、用户绕行距离最小为目标的多场景快充站选址定容优化模型,通过多种线性化手段处理指数等式均衡约束,将优化问题转换为混合整数线性规划以高效求解;最后,以33节点交通系统为例对快充站规划结果进行分析,验证所提模型和线性近似方法的合理性。

中国零碳商用车市场渗透率建模:以重型长途牵引货车为例

商用车碳减排已经成为我国道路交通减碳的关键瓶颈,新能源商用车被视作重型商用车减碳的重要途径,但是新能源商用车的市场渗透率远低于其他车辆部门;但与此同时,现阶段新能源零碳商用车的发展还存在着应用场景复杂、技术路径多样化、同时成本较高的显著的瓶颈。本研究构建了基于新能源汽车总拥有成本(total cost of ownership,TCO)、使用便利性等因素的多元Logit离散选择模型——零碳商用车市场演进模型(discrete choice-based market evolution of green truck model,DC-MEGT),使用自下向上的方法计算TCO,并将车辆使用便利性使用补能时间成本进行货币化量化,构建综合效用函数对纯电动车、燃料电池汽车及零碳燃料等不同动力类型从目前到2060年的市场渗透率演进情况进行预测分析。研究以重型长途牵引场景为例进行分析,结果表明2060年主要的技术路径包括燃料电池汽车、纯电动车、天然气及柴油车,占比分别为48%、28%、12%和10%。政策推广、技术进步、商业模式等因素的不确定性会引发纯电动车和燃料电池汽车2060年市场份额17%~19%的波动。

Combined model of trip mode and destination

This paper analyzes the characteristics of the destination distribution of trips and proposes a stratified sampling strategy for travel mode choice.The stratified sampling strategy can reduce the size of the alternative set;thus,the computation burden of simulation is decreased.Using the stratified sampling strategy,a combined choice model of the trip mode and destination is developed based on the Bayesian theory.Simulations are carried out to verify the proposed model.The results show that the combined choice model of the trip mode and destination can efficiently simulate travelers＇ choice behaviors.Furthermore,the forecasting accuracy of the combined choice model is higher than the one of the gravity model.Therefore,the proposed model is a powerful tool with which to analyze travelers＇ behaviors in selecting the trip mode.

基于离散选择实验的泸州市居民HIV检测偏好研究

目的了解泸州市居民主动参加人类免疫缺陷病毒(HIV)检测情况及检测偏好,分析检测偏好的主要驱动因素,为制定促进该人群主动进行HIV检测的策略提供参考依据。方法采用多阶段分层随机整群抽样的方法,抽取泸州市7个城区和7个乡镇的2279名居民为研究对象进行问卷调查。应用离散选择实验、联合分析的方法编制HIV检测偏好的测量内容、分析泸州市居民HIV检测偏好及其属性的特征,并通过χ^(2)检验与混合logit模型分析HIV检测偏好的驱动因素。结果泸州市居民中主动进行HIV检测的比例为26.20%,检测地点(医疗卫生机构)、额外服务(免费提供)与检测费用(收费)是居民主要的检测偏好,检测地点(β=-0.313,95%CI:-0.587~-0.039,P<0.05)、检测样本(β=-0.422,95%CI:-0.657~-0.187,P<0.05)、检测费用(β=-0.353,95%CI:-0.570~-0.137,P<0.05)、检测时间(β=-0.203,95%CI:-0.404~-0.003,P<0.05)是检测偏好的驱动因素。结论泸州市居民主动进行HIV检测的比例较低,应针对不同人群的检测偏好制定个性化的检测策略,以提高HIV检测的覆盖率和效果。

不同场景下出行链模式与充电选择的影响关系

为了对比研究不同场景下电动汽车用户充电需求产生过程的异质性,选定工作日、非工作日为对象,采用递归联立方程双变量Probit模型分别建模研究两种环境下出行链模式(出行结构)与充电选择之间的影响关系:一种是出行链模式决策影响充电选择,另一种是充电选择影响出行链模式选择。利用出行链及充电选择调查数据对模型进行标定与验证,数据结果表明:工作日、非工作日场景下,模型的拟合结果存在差异,用户出行链模式与充电选择之间的主导关系不同,电动汽车充电需求的产生机理存在异质性。工作日中,出行链决策影响充电选择,表明用户的出行决策先于充电决策,充电需求的产生依附于出行链和活动计划。这种模式下,应充分考虑居民的出行习惯,结合出行目的、停驻点类型、出行时间、车辆续航情况等关键属性进行城市充电需求的预测和调度;而非工作日中,充电选择影响出行链决策,表明用户的充电决策先于出行决策,此时充电需求则较少受到出行计划的影响,应结合用户的个人社会经济属性、车辆属性、充电桩拥有情况等对充电偏好进行分析,依据充电选择情况进行需求预测和调度。因此,工作日、非工作日下存在两类充电需求产生机制,研究结果对城市充电需求预测及调度具有一定的指导意义。

景区预约出行服务对旅游出行方式选择影响研究

基于南京市假日热门景区的调查数据,通过建立潜在类别模型和景区出行方式决策模型,分析预约出行服务对不同出行偏好游客的出行方式转移意向。结果表明按出行偏好游客可分为4类;职业、出行偏好显著影响游客出行方式转变;基于游客需求偏好的预约出行服务能缓解交通资源供需失衡问题。研究认为,景区通过优化预约变更次数、提前截止门票预约时间及加大公交、网约/出租车优惠和吸引力,可提升景区周边交通服务水平。

Pricing Models in Marketing Research

Pricing a product is one of the most important decisions an organization can make. Marketing research has developed several different approaches to price optimization. They include direct methods such as estimation of willingness to pay, indirect methods such as Gabor-Granger and van Westendorp techniques, and product/price mix methods such as various discrete choice models. All of them are widely used in practical marketing research for evaluation of optimal prices for different products and product innovations. This work describes and compares several main of these approaches.

商业综合体消费者行为决策机制及空间优化模拟研究

商业综合体的业态组合及空间布局呈现同质化趋势,难以与电商网络背景下消费者体验式随机消费行为相匹配。研究通过分析商业综合体内消费者行为的影响因素建立离散选择模型,并运用多智能体仿真建模技术模拟空间优化策略的实施效果。针对不同类型消费者存在的消费偏好差异,通过提升品牌知名度、调整业态布局,可以有效改善商业空间的客流分布,为商业动态运营提供可视化的决策工具。

货运价格转换及基于揭示偏好数据的公铁竞争模型

为深入推进货物运输“公转铁”,铁路运输企业需掌握竞争方式同口径可比价格,并充分理解货运方式选择行为。本文提出一种不同装载和运输方式之间的运价转换方法,解决了采用RP(Revealed Preference)数据进行离散选择建模时备选项属性数据缺失的问题。通过改进的PPS(Probability Proportionate to Size Sampling)方法,有效组合多源RP数据,构建货运方式选择行为模型。结果表明,模型能正确预测90%以上的观测值。轻货的VOT(Value of Time)相比重货更高。价格弹性的推导和计算表明,提高公路价格比降低铁路价格能使铁路分担率有更大的提升,降低当前铁路价格可以增加运输收入。当铁路价格下降到收入最大化目标的最优定价点时,不仅会带来铁路分担率、运量和收入的显著增加,还有望获得一定的碳减排效益。

考虑动态潜计划的巡游停车行为分析

为了定量分析驾驶员在停车巡游过程中的微观动态行为决策的内在机理,从动态潜计划理论角度构建了双层结构(停车场选择(计划的选择)和巡游路线选择(基于计划的行动选择))的停车巡游行为分析框架。依托行为决策框架分析了巡游行为的影响因素,并建立了基于动态潜计划的停车巡游行为模型。通过停车行为问卷调查所获数据进行了模型参数标定与检验,并与一般MNL模型作了比较分析,研究了各影响因素在停车巡游动态决策过程中的作用联系。结果表明:依据动态潜计划考虑的影响因素(过去计划、过去行动和当前计划)对停车巡游行为具有显著性影响,对于不同的停车计划,巡游路线的选择行为存在较大的差异性;直接影响计划选择的显著性因素为性别、职业、车辆属性、停车费用、停车时间、便捷程度、停车后步行距离、过去计划、过去行动,直接影响巡游路线选择的显著性因素为性别、驾龄、职业、当前计划、道路环境、车辆属性和信息获取方式;过去计划、过去行动的负面经历最终会影响下一阶段计划和行动的选择;超出忍受的停车后步行距离和停车排队时间,以及巡游路线道路等级、环境美观程度是使驾驶员由路外停车转变为路内停车的显著性影响因素;该模型具有较好的精度和拟合效果,在揭示停车巡游行为动态变化规律上具有较强的解释能力。