Search

• Created a TB-ResNet framework to synergize theory- and data-driven methods. • Exemplified the synergy by combing discrete choice models and deep neural networks. • Adapted TB-ResNets for multinomial logit, risk, and time preferences. • Empirically tested the TB-ResNets for travel behavior analysis. • Illustrated TB-ResNets' strength in prediction, interpretation, and robustness. Researchers often treat data-driven and theory-driven models as two disparate or even conflicting methods in travel behavior analysis. However, the two methods are highly complementary because data-driven methods are more predictive but less interpretable and robust, while theory-driven methods are more interpretable and robust but less predictive. Using their complementary nature, this study designs a theory-based residual neural network (TB-ResNet) framework, which synergizes discrete choice models (DCMs) and deep neural networks (DNNs) based on their shared utility interpretation. The TB-ResNet framework is simple, as it uses a (δ , 1- δ) weighting to take advantage of DCMs' simplicity and DNNs' richness, and to prevent underfitting from the DCMs and overfitting from the DNNs. This framework is also flexible: three instances of TB-ResNets are designed based on multinomial logit model (MNL-ResNets), prospect theory (PT-ResNets), and hyperbolic discounting (HD-ResNets), which are tested on three data sets. Compared to pure DCMs, the TB-ResNets provide greater prediction accuracy and reveal a richer set of behavioral mechanisms owing to the utility function augmented by the DNN component in the TB-ResNets. Compared to pure DNNs, the TB-ResNets can modestly improve prediction and significantly improve interpretation and robustness, because the DCM component in the TB-ResNets stabilizes the utility functions and input gradients. Overall, this study demonstrates that it is both feasible and desirable to synergize DCMs and DNNs by combining their utility specifications under a TB-ResNet framework. Although some limitations remain, this TB-ResNet framework is an important first step to create mutual benefits between DCMs and DNNs for travel behavior modeling, with joint improvement in prediction, interpretation, and robustness. [ABSTRACT FROM AUTHOR]

Classical demand modeling analyzes travel behavior using only low-dimensional numeric data (i.e. sociodemographics and travel attributes) but not high-dimensional urban imagery. However, travel behavior depends on the factors represented by both numeric data and urban imagery, thus necessitating a synergetic framework to combine them. This study creates a theoretical framework of deep hybrid models consisting of a mixing operator and a behavioral predictor, thus integrating the numeric and imagery data for travel behavior analysis. Empirically, this framework is applied to analyze travel mode choice using the Chicago MyDailyTravel Survey as the numeric inputs and the satellite images as the imagery inputs. We found that deep hybrid models significantly outperform both classical demand models and deep learning models in predicting aggregate and disaggregate travel behavior. The deep hybrid models can reveal spatial clusters with meaningful sociodemographic associations in the latent space. The models can also generate new satellite images that do not exist in reality and compute the corresponding economic information, such as substitution patterns and social welfare. Overall, the deep hybrid models demonstrate the complementarity between the low-dimensional numeric and high-dimensional imagery data and between the traditional demand modeling and recent deep learning. They enrich the family of hybrid demand models by using deep architecture as the latent space and enabling researchers to conduct associative analysis for sociodemographics, travel decisions, and generated satellite imagery. Future research could address the limitations in interpretability, robustness, and transferability, and propose new methods to further enrich the deep hybrid models. • Designing deep hybrid models to integrate numeric and imagery data. • Deep hybrid models outperform demand models and deep learning in demand prediction. • Deep hybrid models have high-dimensional latent space for image processing and generation. • The latent space is spatially and socially meaningful. • Deep hybrid models can generate new urban imagery and derive economic information. [ABSTRACT FROM AUTHOR]

It is an enduring question how to combine revealed preference (RP) and stated preference (SP) data to analyze individual choices. While the nested logit (NL) model is the classical way to address the question, this study presents multitask learning deep neural networks (MTLDNNs) as an alternative framework, and discusses its theoretical foundation, empirical performance, and behavioral intuition. We first demonstrate that the MTLDNNs are theoretically more general than the NL models because of MTLDNNs' automatic feature learning, flexible regularizations, and diverse architectures. By analyzing the adoption of autonomous vehicles (AVs), we illustrate that the MTLDNNs outperform the NL models in terms of prediction accuracy but underperform in terms of cross-entropy losses. To interpret the MTLDNNs, we compute the elasticities and visualize the relationship between choice probabilities and input variables. The MTLDNNs reveal that AVs mainly substitute driving and ride hailing, and that the variables specific to AVs are more important than the socio-economic variables in determining AV adoption. Overall, this work demonstrates that MTLDNNs are theoretically appealing in leveraging the information shared by RP and SP and capable of revealing meaningful behavioral patterns, although its performance gain over the classical NL model is still limited. To improve upon this work, future studies can investigate the inconsistency between prediction accuracy and cross-entropy losses, novel MTLDNN architectures, regularization design for the RP-SP question, MTLDNN applications to other choice scenarios, and deeper theoretical connections between choice models and the MTLDNN framework. [ABSTRACT FROM AUTHOR]

Deep neural networks (DNNs) have been increasingly applied in travel demand modeling because of their automatic feature learning, high predictive performance, and economic interpretability. Nevertheless, DNNs frequently present behaviorally irregular patterns, significantly limiting their practical potentials and theoretical validity in travel behavior modeling. This study proposes strong and weak behavioral regularities as novel metrics to evaluate the monotonicity of individual demand functions (known as the "law of demand"), and further designs a constrained optimization framework with six gradient regularizers to enhance DNNs' behavioral regularity. The empirical benefits of this framework are illustrated by applying these regularizers to travel survey data from Chicago and London, which enables us to examine the trade-off between predictive power and behavioral regularity for large versus small sample scenarios and in-domain versus out-of-domain generalizations. The results demonstrate that, unlike models with strong behavioral foundations such as the multinomial logit, the benchmark DNNs cannot guarantee behavioral regularity. However, after applying gradient regularization, we increase DNNs' behavioral regularity by around 6 percentage points while retaining their relatively high predictive power. In the small sample scenario, gradient regularization is more effective than in the large sample scenario, simultaneously improving behavioral regularity by about 20 percentage points and log-likelihood by around 1.7%. Compared with the in-domain generalization of DNNs, gradient regularization works more effectively in out-of-domain generalization: it drastically improves the behavioral regularity of poorly performing benchmark DNNs by around 65 percentage points, highlighting the criticality of behavioral regularization for improving model transferability and applications in forecasting. Moreover, the proposed optimization framework is applicable to other neural network–based choice models such as TasteNets. Future studies could use behavioral regularity as a metric along with log-likelihood, prediction accuracy, and F 1 score when evaluating travel demand models, and investigate other methods to further enhance behavioral regularity when adopting complex machine learning models. • Designing behavioral regularity metrics using the law of demand. • Using behavioral regularity to evaluate deep learning in choice analysis. • Instilling behavioral regularity into deep learning through gradient regularization. • Empirically, gradient regularization improves behavioral regularity of deep learning. • Gradient regularization also enhances prediction in small samples. [ABSTRACT FROM AUTHOR]

Traffic crashes present substantial challenges to human safety and socio-economic development in urban areas. Developing a reliable and responsible traffic crash prediction model is crucial to address growing public safety concerns and improve the safety of urban mobility systems. Traditional methods face limitations at fine spatiotemporal scales due to the sporadic nature of high-risk crashes and the predominance of non-crash characteristics. Furthermore, while most current models show promising occurrence prediction, they overlook the uncertainties arising from the inherent nature of crashes, and then fail to adequately map the hierarchical ranking of crash risk values for more precise insights. To address these issues, we introduce the S patio t emporal Z ero- I nflated T wee d ie G raph N eural N etworks (STZITD-GNN), the first uncertainty-aware probabilistic graph deep learning model in road-level daily-basis traffic crash prediction for multi-steps. Our model combines the interpretability of the statistical Tweedie family with the predictive power of graph neural networks, excelling in predicting a comprehensive range of crash risks. The decoder employs a compound Tweedie model, handling the non-Gaussian distribution inherent in crash data, with a zero-inflated component for accurately identifying non-crash cases and low-risk roads. The model accurately predicts and differentiates between high-risk, low-risk, and no-risk scenarios, providing a holistic view of road safety that accounts for the full spectrum of probability and severity of crashes. Empirical tests using real-world traffic data from London, UK, demonstrate that the STZITD-GNN surpasses other baseline models across multiple benchmarks, including a reduction in regression error of up to 34.60% in point estimation metrics and an improvement of above 47% in interval-based uncertainty metrics. • Introduces a novel probabilistic graph deep learning model (STZITD-GNN) for road-level crash risk assessment. • Combines graph neural networks with a four-parameter statistical distribution to effectively handle zero-inflated traffic crash data. • Demonstrates superior performance in multi-step road-level crash prediction for both point estimation and uncertainty quantification. • Exhibits stable performance in multi-step predictions and provides insights on deep learning model uncertainties. [ABSTRACT FROM AUTHOR]