Search

The anchor-based nested logit model is suitable to describe route choice behaviors since its abstract framework is consistent with style of travelers processing road network information. But in practice, the traditional definition of anchors is largely related to the properties of road network elements and too many anchors defined greatly increase computational burden. In this study, a data-driven anchor point generation method by community detection was proposed to address these issues. Travel communities are detected considering travel relationship topology by the massive individual trip data; and the frequently used bridges, expressways, and arterial roads in one community are identified as an anchor point which is in accordance with the mental representation of travelers’ route choice. The anchor points are employed to construct the nests in the route choice model to capture the choice correlation of travelers who pass through the same anchor points; and the small number of travel communities means the nests can be significantly reduced in number comparing with the traditional anchor-based models so that the computing burden is much less, while the ability to capture the correlation of routes is still remained. A case study is carried out for Guangzhou City, and the results suggest that the proposed anchor point based model obtains satisfying results in goodness-of-fit and forecasting; in addition, the computation time counts about only one-tenth that of the traditional anchor-based model, making it suitable for route choice analysis in the large road networks. Moreover, some practical implications are drawn for traffic management from the empirical study.

The ageing population has become a global problem in which enhanced understanding on their activity-travel patterns is needed. In this paper, an analysis of retired and dual-earner couples is conducted to investigate how retirement would change their activity time use and patterns. In particular, intra-household interactions are considered, to explore the interdependencies among household members’ choices, social-demographics and travel behaviours. Household survey data from Hong Kong are employed. Results show that retirement would substantially increase joint participations and durations in various out-of-home activities. In addition, the importance of walkability is emphasised for retired couples in a mixed-land-use and transit-dependent city, and a potential social exclusion issue is identified for the low-income retired population. Scenarios analyses including changes of built environment and lifestyles (e.g., telecommuting, online shopping and food delivery) are conducted, to investigate how couples would reallocate the saved travel time. In summary, this paper highlights the importance of considering the group decision mechanism in a household for activity generation and travel demand forecasting. It sheds light on policies to improve quality-of-life for couples before and after the retirement.

The application of travel demand models to transportation planning has triggered great interests in issues that potentially improve the accuracy of model forecasts. These forecasts, however, are su...

The analysis of travel mode choice is important in transportation planning and policy-making in order to understand and forecast travel demands. Research in the field of machine learning has been exploring the use of random forest as a framework within which many traffic and transport problems can be investigated. The random forest (RF) is a powerful method for constructing an ensemble of random decision trees. It de-correlates the decision trees in the ensemble via randomization that leads to an improvement of forecasting and reduces the variance when averaged over the trees. However, the usefulness of RF for travel mode choice behavior remains largely unexplored. This paper proposes a robust random forest method to analyze travel mode choices for examining the prediction capability and model interpretability. Using the travel diary data from Nanjing, China in 2013, enriched with variables on the built environment, the effects of different model parameters on the prediction performance are investigated. The comparison results show that the random forest method performs significantly better in travel mode choice prediction for higher accuracy and less computation cost. In addition, the proposed method estimates the relative importance of explanatory variables and how they relate to mode choices. This is fundamental for a better understanding and effective modeling of people’s travel behavior.

To accommodate the diverse users demands for consumer products, enterprises need to design and develop different lines of products according to different groups of users. Dynamic internet data, inc...

Drivers' route choice model is essential in transportation software such as navigation, fleet management, and simulation, where the random utility models (RUM) have dominated for years. The authors investigate here whether machine learning (ML) models could be applied into this field, and whether these approaches outperform the traditional models in goodness-of-fit and prediction. The application framework and data structure are proposed, where the challenging problems lie in: (i) to pool data from multiple origin-destination pairs; and (ii) to interpret results for behaviour analysis. All RUM and ML models are applied in a real network. Results suggest that the random forest, one of the ML models, has satisfying performances with acceptable computation time, making it suitable for large network and real-time analysis. This study shows that the ML models can be adopted for behaviour analysis, such as to prioritise the importance of variables, compute the elasticity, and forecast for scenarios. Future directions on combining the RUM and ML models are discussed.

Purpose The purpose of this paper is to propose an effective and economical management platform to realize real-time tracking and tracing for prepackaged food supply chain based on Internet of Things (IoT) technologies, and finally ensure a benign and safe food consumption environment. Design/methodology/approach Following service-oriented architecture, a flexible layered architecture of tracking and tracing platform for prepackaged food is developed. Besides, to reduce the implementation cost while realizing fine-grained tracking and tracing, an integrated solution of using both the QR code and radio-frequency identification (RFID) tag is proposed. Furthermore, Extensible Markup Language (XML) is adopted to facilitate the information sharing among applications and stakeholders. Findings The validity of the platform has been evaluated through a case study. First, the proposed platform is proved highly effective on realizing prepackaged food tracking and tracing throughout its supply chain, and can benefit all the stakeholders involved. Second, the integration of the QR code and RFID technologies is proved to be economical and could well ensure the real-time data collection. Third, the XML-based method is efficient to realize information sharing during the whole process. Originality/value The contributions of this paper lie in three aspects. First, the technical architecture of IoT-based tracking and tracing platform is developed. It could realize fine-grained tracking and tracing and could be flexible to adapt in many other areas. Second, the solution of integrating the QR code and RFID technologies is proposed, which could greatly decrease the cost of adopting the platform. Third, this platform enables the information sharing among all the involved stakeholders, which will further facilitate their cooperation on guaranteeing the quality and safety of prepackaged food.

Nowadays, the society is paying more and more attention to the issues that plague the environment. There is a growing need for enterprises to design and manufacture environmentally friendly products, in order to take up the responsibility of sustainable development of the society. The entire product life cycle consists of design, manufacturing, usage and recycling phases. These phases involve various economic and environmental factors that affect the recyclability of products. Previous researchers mainly focused on the single phase of the product’s life cycle. Moreover, they did not take into account the dynamic change of the recyclable material’s cost. In this paper, a time-series forecasting methodology is proposed to evaluate the product’s recyclability at the product design phase. It considers various economic and environmental factors of different stages of the product’s life cycle. In addition, a time-series forecasting method is utilised for predicting the cost of the recycled material at the produ...

A novel concept is presented to capture route choice behaviours and to account for the correlation of routes in the logit model. The issue that route choice models are easily affected when irrelevant alternatives are included in the choice set is tackled. The concept of equivalent impedance is presented to simplify and aggregate a set of links based on the idea that people tend to remember and process road network information at an abstract level. Then, the equivalent impedance is utilized to derive a correction term for the utility of a multinomial logit model in which the advantages of a closed-form structure and easy computation remain unchanged. The results from the numerical examples suggest that the proposed model obtains reasonable results and provides more stable predictions than comparable models when the composition of the choice set changes. An application in a real urban network with GPS data is presented, and estimation results suggest that the new model is practical due to its robustness.

Automotive aftermarket (e.g. repair and maintenance) is one of the most lucrative business in the entire automotive industry chain. However, with the fierce competition recently, many service providers (also known as 4S shops) are facing a problem of customer churn. It would be most valuable to investigate the determinants of customers' loyalty. Ten years' data of a 4S shop is analyzed where customers' social-demographics, maintenance patterns and habits, car characteristics, repair types, fees and discounts, etc. are available, and machine learning approaches are employed. This paper investigates the causes for the customer churn in 4S shops, and proposes several solutions to improve the management and operation of 4S shops. Results from this analysis shed light on customers' loyalty behaviors and automotive industry customer relationship management.

A subpath-based methodology is proposed to capture the travellers’ route choice behaviours and their perceptual correlation of routes, because the original link-based style may not be suitable in application: (1) travellers do not process road network information and construct the chosen route by a link-by-link style; (2) observations from questionnaires and GPS data, however, are not always link-specific. Subpaths are defined as important portions of the route, such as major roads and landmarks. The cross-nested Logit (CNL) structure is used for its tractable closed-form and its capability to explicitly capture the routes correlation. Nests represent subpaths other than links so that the number of nests is significantly reduced. Moreover, the proposed method simplifies the original link-based CNL model; therefore, it alleviates the estimation and computation difficulties. The estimation and forecast validation with real data are presented, and the results suggest that the new method is practical.

Repair and maintenance services are among the most lucrative aspects of the entire automobile business chain. However, in the context of fierce competition, customer churns have led to the bankruptcy of several 4S (sales, spare parts, services, and surveys) shops. In this regard, a six-year dataset is utilized to study customer behaviors to aid managers identify and retain valuable but potential customer churn through a customized retention solution. First, we define the absence and presence behaviors of customers and thereafter generate absence data according to customer habits; this makes it possible to treat the customer absence prediction problem as a classification problem. Second, the repeated absence and presence behaviors of customers are considered as a whole from a lifecycle perspective. A modified recurrent neural network (RNN-2L) is proposed; it is more efficient and reasonable in structure compared with traditional RNN. The time-invariant customer features and the sequential lifecycle features are handled separately; this provides a more sensible specification of the RNN structure from a behavioral interpretation perspective. Third, a customized retention solution is proposed. By comparing the proposed model with those that are conventional, it is found that the former outperforms the latter in terms of area under the curve (AUC), confusion matrix, and amount of time consumed. The proposed customized retention solution can achieve significant profit increase. This paper not only elucidates the customer relationship management in the automobile aftermarket (where the absence and presence behaviors are infrequently considered), but also presents an efficient solution to increase the predictive power of conventional machine learning models. The latter is achieved by considering behavioral and business perspectives.

A Logit-based route choice model is proposed to address the overlapping and scaling problems in the traditional multinomial Logit model. The nonoverlapping links are defined as a subnetwork, and its equivalent impedance is explicitly calculated in order to simply network analyzing. The overlapping links are repeatedly merged into subnetworks with Logit-based equivalent travel costs. The choice set at each intersection comprises only the virtual equivalent route without overlapping. In order to capture heterogeneity in perception errors of different sizes of networks, different scale parameters are assigned to subnetworks and they are linked to the topological relationships to avoid estimation burden. The proposed model provides an alternative method to model the stochastic route choice behaviors without the overlapping and scaling problems, and it still maintains the simple and closed-form expression from the MNL model. A link-based loading algorithm based on Dial’s algorithm is proposed to obviate route enumeration and it is suitable to be applied on large-scale networks. Finally a comparison between the proposed model and other route choice models is given by numerical examples.

An improved paired combinatorial Logit route choice model with Probit-based equivalent route impedance is proposed to simplify the calculation and resolve the homoscedasticity problem of the Logit model. The model comprises a two-level structure: the lower level employs a binary Probit choice model to address the problem of route overlapping. Clark's approximation for normal distribution is employed to compute the equivalent impendence for each route pair in the upper level. The homoscedasticity problem is resolved through the introduction of the normal distribution. The upper level is a classical multinomial Logit model with relative impedance that reduces the defect of variance-homogeneity in the upper level. The probability of route selection is determined by marginal and conditional probabilities. The proposed model combines the advantages of both Probit and Logit models with close-form formulations, which can be easily calculated. Two numerical tests, which were performed on the well-known examples, indicate that the proposed model produces more reasonable and stable results.

In the paper, the modified Logit route choice models with relative impedance are presented to resolve the fixed variances problem of the traditional logit model. Therefore, the variances of perception error are relative to the route impedances. The classical variations of Logit models employing relative impedances are deduced. Three numerical tests show that the proposed models are more reasonable and robust than the traditional ones, and the modified PCL model with Probit equivalent impedances is the closest to theoretical results. The revised models have clear theoretical background, perform more rationally than traditional ones and still retain the advantage of easy computation, which makes them flexible in research and practical application.

The selection of fresh product suppliers is a multi-criteria decision making (MCDM) problem with great significant and application value. This requires trade-offs between multiple criteria to prove its ambiguity and uncertainty. Therefore, a novel two-stage fuzzy integrated MCDM method to select suitable suppliers is employed. In the first stage, two collective relationship matrixes are constructed by quality function development (QFD), and relationships among customer requirements (CRs), company strategies (CSs) as well as selection criteria are considered separately in the two matrixes. Subjective criteria weights are obtained by fuzzy best-worst method (BWM) appropriately. In the second stage, the objective criteria weights are obtained using Shannon&rsquo, s entropy method, and the fuzzy multi-objective optimization by ratio analysis plus the full multiplicative form (MULTIMOORA) is applied to rank suppliers. Finally, an application case is applied to prove the feasibility of the proposed method. These conclusions can help companies improve their CSs and increase their market competitiveness.

A closed-form mixed Logit approach is proposed to model the stochastic route choice behaviours. It combines both the advantages of Probit and Logit to provide a flexible form in alternatives correlation and a tractable form in expression; besides, the heterogeneity in alternative variance can also be addressed. Paths are compared by pairs where the superiority of the binary Probit can be fully used. The Probit-based aggregation is also used for a nested Logit structure. Case studies on both numerical and empirical examples demonstrate that the new method is valid and practical. This paper thus provides an operational solution to incorporate the normal distribution in route choice with an analytical expression.

The high traffic volume in Chinese universities brings issues including accessibility and safety, which requires careful evaluation of campus-traffic with a quantitative and qualitative method. A multilevel index system based on attribute mathematical recognition is proposed. The upper-level includes five indexes, namely efficiency, accessibility, comfort, safety and traffic calming, which are further divided into the lower-level indexes. The indexes' weights are determined by variation coefficient method and the campus traffic level was identified by the confidence rules to avoid unreasonable estimation. A case study of Sun Yat-sen University East Campus is presented, whose travel environment was just passable in 2007 but greatly improved since 2009 after the road reconstruction and traffic regulations suggested by the authors. The evaluation accords with the opinion poll and the method is practical, rational and suitable for comparison.

We present an operational estimation procedure for the estimation of route choice multivariate extreme value (MEV) models based on sampling of alternatives. The procedure builds on the state-of-the-art literature, and in particular on recent methodological developments proposed by Flotterod and Bierlaire (2013) and Guevara and Ben-Akiva (2013b). Case studies on both synthetic data and a real network demonstrate that the new method is valid and practical.