Travel Time Estimation (TTE) stands as a cornerstone of efficient transportation systems. However, the critical imperative of privacy preservation within the TTE context remains notably underexplored. This gap underscores the pressing necessity for innovative solutions that prioritize the safeguarding of users' geo-privacy, particularly in light of the expanding prevalence of data-driven TTE algorithms. In this paper, a novel privacy-preserving TTE framework, CRATE, is proposed to ensure comprehensive privacy preservation for TTE without compromising service quality. CRATE achieves this objective by identifying random routes within a transportation network that yield identical travel times to the actual, privacy-rich route. This is accomplished through exploiting the embedding representations for road segments and routes, followed by the development of a highly efficient heuristic for random route generation. Furthermore, a travel time aggregation and calibration model is devised to enhance estimation accuracy while upholding user privacy. Case studies conducted on three real-world vehicular trajectory datasets demonstrate that CRATE attains comparable estimation accuracy to state-of-the-art non-privacy-preserving TTE algorithms while maintaining strict privacy protection. Additionally, CRATE's efficiency is showcased through deployment on both high- and low-end mobile handsets spanning the past decade.

In the evolving field of urban development, precise traffic prediction is essential for optimizing traffic and mitigating congestion. While traditional graph learning-based models effectively exploit complex spatial-temporal correlations, their reliance on trivially generated graph structures or deeply intertwined adjacency learning without supervised loss significantly impedes their efficiency. This paper presents Contrastive Learning of spatial-tEmporal trAffic data Representations (CLEAR) framework, a comprehensive approach to spatial-temporal traffic data representation learning aimed at enhancing the accuracy of traffic predictions. Employing self-supervised contrastive learning, CLEAR strategically extracts discriminative embeddings from both traffic time-series and graph-structured data. The framework applies weak and strong data augmentations to facilitate subsequent exploitations of intrinsic spatial-temporal correlations that are critical for accurate prediction. Additionally, CLEAR incorporates advanced representation learning models that transmute these dynamics into compact, semantic-rich embeddings, thereby elevating downstream models’ prediction accuracy. By integrating with existing traffic predictors, CLEAR boosts predicting performance and accelerates the training process by effectively decoupling adjacency learning from correlation learning. Comprehensive experiments validate that CLEAR can robustly enhance the capabilities of existing graph learning-based traffic predictors and provide superior traffic predictions with a straightforward representation decoder. This investigation highlights the potential of contrastive representation learning in developing robust traffic data representations for traffic prediction.

Generating trajectory data is among promising solutions to addressing privacy concerns, collection costs, and proprietary restrictions usually associated with human mobility analyses. However, existing trajectory generation methods are still in their infancy due to the inherent diversity and unpredictability of human activities, grappling with issues such as fidelity, flexibility, and generalizability. To overcome these obstacles, we propose ControlTraj, a Controllable Trajectory generation framework with the topology-constrained diffusion model. Distinct from prior approaches, ControlTraj utilizes a diffusion model to generate high-fidelity trajectories while integrating the structural constraints of road network topology to guide the geographical outcomes. Specifically, we develop a novel road segment autoencoder to extract fine-grained road segment embedding. The encoded features, along with trip attributes, are subsequently merged into the proposed geographic denoising UNet architecture, named GeoUNet, to synthesize geographic trajectories from white noise. Through experimentation across three real-world data settings, ControlTraj demonstrates its ability to produce human-directed, high-fidelity trajectory generation with adaptability to unexplored geographical contexts.

Graph-based deep learning models are becoming prevalent for data-driven traffic prediction in the past years, due to their competence in exploiting the non-euclidean spatial-temporal traffic data. Nonetheless, these models are approaching a limit where drastically increasing model complexity in terms of trainable parameters cannot notably improve the prediction accuracy. Furthermore, the diversity of transportation networks requires traffic predictors to be scalable to various data sizes and quantities, and ever-changing traffic dynamics also call for capacity sustainability. To this end, we propose a novel adaptive deep learning scheme for boosting graph-based traffic predictor performance. The proposed scheme utilizes domain knowledge to decompose the traffic prediction task into sub-tasks, each of which is handled by deep models with low complexity and training difficulty. Further, a stream learning algorithm based on the empirical Fisher information loss is devised to enable predictors to incrementally learn from new data without re-training from scratch. Comprehensive case studies on five real-world traffic datasets indicate outstanding performance improvement of the proposed scheme when equipped to six state-of-the-art predictors. Additionally, the scheme also provides impressive autoregressive long-term predictions and incremental learning efficacy with traffic data streams.

Uncertainty is an essential consideration for time series forecasting tasks. In this work, we specifically focus on quantifying the uncertainty of traffic forecasting. To achieve this, we develop Deep Spatio-Temporal Uncertainty Quantification (DeepSTUQ), which can estimate both aleatoric and epistemic uncertainty. We first leverage a spatio-temporal model to model the complex spatio-temporal correlations of traffic data. Subsequently, two independent sub-neural networks maximizing the heterogeneous log-likelihood are developed to estimate aleatoric uncertainty. For estimating epistemic uncertainty, we combine the merits of variational inference and deep ensembling by integrating the Monte Carlo dropout and the Adaptive Weight Averaging re-training methods, respectively. Finally, we propose a post-processing calibration approach based on Temperature Scaling, which improves the model’s generalization ability to estimate uncertainty. Extensive experiments are conducted on four public datasets, and the empirical results suggest that the proposed method outperforms state-of-the-art methods in terms of both point prediction and uncertainty quantification.

Estimation of road link travel time serves a critical role in intelligent transportation operation and management. Due to the uncertainty nature contributed by the volatile traffic, travel time estimates are better described by probability distributions than deterministic models. Existing travel time distribution estimation approaches are mostly based on predefined probability distributions. Other approaches, while relaxing the constraint, fail to utilize the topological information and are data-inefficient. In this paper, we propose a novel Bayesian and geometric deep learning-based approach to estimate the travel time distributions of road links within citywide transportation networks based on vehicular GPS trajectories. Particularly, historical or real-time trajectories are first pre-processed to construct partial travel time maps, which are input into a tailor-made Bayesian graph autoencoder to reconstruct multiple complete travel time maps. We further adopt an auxiliary neural network to facilitate the parameter training of the proposed approach following adversarial training principles. To evaluate the proposed approach, we employ a real-world vehicular trajectory dataset in a series of comprehensive case studies. The empirical results indicate that the proposed approach outperforms the best-performing state-of-the-art baseline with an approximately 10% Kullback-Leibler divergence reduction.

With the increasing awareness of the harmful impacts of urban air pollution, air quality monitoring stations have been deployed in many metropolitan areas. These stations provide air quality data to the public. However, due to sampling device failures and data processing errors, missing data in air quality measurements is common. Data integrity becomes a critical challenge when such data are employed for public services. In this paper, we investigate the mathematical property of air quality measurements, and attempt to recover the missing data. First, we empirically study the low rank property of these measurements. Second, we formulate the low rank matrix completion (LRMC) optimization problem to reconstruct the missing air quality data. The problem is transformed using duality theory, and singular value thresholding (SVT) is employed to develop sub-optimal solutions. Third, to evaluate the performance of our methodology, we conduct a series of case studies including different types of missing data patterns. The simulation results demonstrate that the proposed SVT methodology can effectively recover missing air quality data, and outperform the existing Interpolation. Finally, we investigate the parameter sensitivity of SVT. Our study can serve as a guideline for missing data recovery in the real world.