# DSETA: A Dual-Stage Continual Learning Framework for Travel Time Prediction in Dynamic Traffic Environments

arXiv: https://arxiv.org/abs/2608.00402 · HTML: https://mlsanity.com/paper/2608.00402/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 47 references that are in the map.
- First posted on arXiv 1 Aug 2026.

## Builds on (most cited first, 19 of 19)

- [Continual Lifelong Learning with Neural Networks: A Review](https://mlsanity.com/paper/1802.07569/) (2018)
- [Graph WaveNet for Deep Spatial-Temporal Graph Modeling](https://mlsanity.com/paper/1906.00121/) (2019)
- [DeepFM: A Factorization-Machine based Neural Network for CTR Prediction](https://mlsanity.com/paper/1703.04247/) (2017)
- [Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting](https://mlsanity.com/paper/2007.02842/) (2020)
- [On the Effectiveness of Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2211.15583/) (2022)
- [PDFormer: Propagation Delay-Aware Dynamic Long-Range Transformer for Traffic Flow Prediction](https://mlsanity.com/paper/2301.07945/) (2023)
- [ETA Prediction with Graph Neural Networks in Google Maps](https://mlsanity.com/paper/2108.11482/) (2021)
- [A Simple Baseline for Travel Time Estimation using Large-Scale Trip Data](https://mlsanity.com/paper/1512.08580/) (2015)
- [Hybrid Spatio-Temporal Graph Convolutional Network: Improving Traffic Prediction with Navigation Data](https://mlsanity.com/paper/2006.12715/) (2020)
- [DeepTravel: a Neural Network Based Travel Time Estimation Model with Auxiliary Supervision](https://mlsanity.com/paper/1802.02147/) (2018)
- [DSLR: Diversity Enhancement and Structure Learning for Rehearsal-based Graph Continual Learning](https://mlsanity.com/paper/2402.13711/) (2024)
- [DuETA: Traffic Congestion Propagation Pattern Modeling via Efficient Graph Learning for ETA Prediction at Baidu Maps](https://mlsanity.com/paper/2208.06979/) (2022)
- [Continual Learning for Smart City: A Survey](https://mlsanity.com/paper/2404.00983/) (2024)
- [DutyTTE: Deciphering Uncertainty in Origin-Destination Travel Time Estimation](https://mlsanity.com/paper/2408.12809/) (2024)
- [Temporal-Aware Evaluation and Learning for Temporal Graph Neural Networks](https://mlsanity.com/paper/2412.07273/) (2024)
- [Predictive Analytics for Enhancing Travel Time Estimation in Navigation Apps of Apple, Google, and Microsoft](https://mlsanity.com/paper/1705.08509/) (2017)
- [TITAN: A Spatiotemporal Feature Learning Framework for Traffic Incident Duration Prediction](https://mlsanity.com/paper/1911.08684/) (2019)
- [GCT-TTE: Graph Convolutional Transformer for Travel Time Estimation](https://mlsanity.com/paper/2306.04324/) (2023)
- [Graph Neural Networks for Travel Distance Estimation and Route Recommendation Under Probabilistic Hazards](https://mlsanity.com/paper/2501.09803/) (2025)

## Abstract (arXiv)

Estimated Time of Arrival (ETA) prediction is a core component of intelligent transportation systems. As traffic congestion patterns become increasingly dynamic in large cities, maintaining high prediction accuracy poses a major challenge for ride-hailing platforms. Existing methods either fail to adapt to irregular traffic patterns and sudden congestion, or suffer from new distributions without disentangling long-term trends from short-term fluctuations, thereby degrading model performance in real-world scenarios. To address this challenge, we propose DSETA, an incrementally updated Dual-Stage ETA prediction framework. Specifically, the continual learning process is divided into textitinter-day and textitintra-day stages. We first design the textitintra-day learning stage, which relies entirely on real-time data to enable dynamic adaptation to short-term traffic patterns caused by events like holidays or accidents. Next, we develop the textitinter-day learning stage, which leverages aggregated historical data from a short time window to capture knowledge of long-term distribution shifts, such as seasonal trends and traffic network evolution. Subsequently, to prevent catastrophic forgetting and preserve knowledge of regular patterns, we explore a textitHistorical Traffic Knowledge Consolidation module. Finally, we validate DSETA's effectiveness and robustness through extensive offline and online experiments conducted on real-world datasets from DiDi's platform. Online A/B tests across three major cities including Beijing, Wuhan, and Xi'an consistently demonstrated performance gains, achieving MAE reductions of 6.62%, 0.73%, and 2.40% respectively. This framework has been successfully deployed in DiDi's production environment, processing hundreds of millions of daily requests and validating its strong performance in industrial applications.