XXLTraffic: Expanding and Extremely Long Traffic Dataset for Ultra-Dynamic Forecasting Challenges

Du Yin,Hao Xue,Arian Prabowo,Shuang Ao,Flora Salim

2024-06-18

Abstract:Traffic forecasting is crucial for smart cities and intelligent transportation initiatives, where deep learning has made significant progress in modeling complex spatio-temporal patterns in recent years. However, current public datasets have limitations in reflecting the ultra-dynamic nature of real-world scenarios, characterized by continuously evolving infrastructures, varying temporal distributions, and temporal gaps due to sensor downtimes or changes in traffic patterns. These limitations inevitably restrict the practical applicability of existing traffic forecasting datasets. To bridge this gap, we present XXLTraffic, the largest available public traffic dataset with the longest timespan and increasing number of sensor nodes over the multiple years observed in the data, curated to support research in ultra-dynamic forecasting. Our benchmark includes both typical time-series forecasting settings with hourly and daily aggregated data and novel configurations that introduce gaps and down-sample the training size to better simulate practical constraints. We anticipate the new XXLTraffic will provide a fresh perspective for the time-series and traffic forecasting communities. It would also offer a robust platform for developing and evaluating models designed to tackle ultra-dynamic and extremely long forecasting problems. Our dataset supplements existing spatio-temporal data resources and leads to new research directions in this domain.

Machine Learning,Artificial Intelligence

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper primarily addresses the following issues: 1. **Limitations of Existing Datasets**: - Existing public traffic datasets have limitations in reflecting the ultra-dynamic characteristics of real-world scenarios. These limitations include continuously evolving infrastructure, changes in temporal distribution, and time intervals caused by sensor downtime or changes in traffic patterns. - These limitations inevitably restrict the practical application of existing traffic prediction datasets. 2. **Need for Ultra-Long Time Span Datasets**: - A new dataset named XXLTraffic is proposed, which is currently the largest public traffic dataset with the longest time span and an increasing number of sensor nodes year by year. This dataset aims to support ultra-dynamic prediction research. - The dataset includes typical time series prediction settings (hourly and daily aggregated data) and introduces time intervals to better simulate real-world constraints. 3. **Challenges of Ultra-Dynamic Prediction**: - The ultra-dynamic challenge includes three key aspects: the continuously evolving state of infrastructure over time, changes in temporal distribution over extremely long observation periods, and spatiotemporal dynamics with time intervals. - These factors require models to adapt to changes in patterns and trends over long time spans and handle inconsistencies and interruptions in the data. By proposing the XXLTraffic dataset, the paper aims to provide new perspectives for the time series and traffic prediction community and offer a robust platform for developing and evaluating models that address ultra-dynamic and ultra-long time prediction problems. Additionally, this dataset complements existing spatiotemporal data resources and guides new research directions in the field.

XXLTraffic: Expanding and Extremely Long Traffic Dataset for Ultra-Dynamic Forecasting Challenges

Urban Traffic Flow Forecasting Based on Memory Time-Series Network.

XTraffic: A Dataset Where Traffic Meets Incidents with Explainability and More

Daily Long-Term Traffic Flow Forecasting Based on a Deep Neural Network

Incorporating Long-term Data in Training Short-term Traffic Prediction Model

Short-Term Traffic Forecasting Using High-Resolution Traffic Data

An adaptive composite time series forecasting model for short-term traffic flow

Multi-scale feature enhanced spatio-temporal learning for traffic flow forecasting

Similarity-based Feature Extraction for Large-scale Sparse Traffic Forecasting

Hybrid Extreme Learning for Reliable Short-Term Traffic Flow Forecasting

An Effective Dynamic Spatio-temporal Framework with Multi-Source Information for Traffic Prediction

Deep transformer-based heterogeneous spatiotemporal graph learning for geographical traffic forecasting

Temporal Super-Resolution Traffic Flow Forecasting Via Continuous-Time Network Dynamics

Distil the informative essence of loop detector data set: Is network-level traffic forecasting hungry for more data?

OpenCity: Open Spatio-Temporal Foundation Models for Traffic Prediction

CALTM: A Context-Aware Long-Term Time-Series Forecasting Model

LSTTN: A Long-Short Term Transformer-based Spatio-temporal Neural Network for Traffic Flow Forecasting

Attention based spatiotemporal graph attention networks for traffic flow forecasting

DuTraffic: Live Traffic Condition Prediction with Trajectory Data and Street Views at Baidu Maps

A Multi-Channel Spatial-Temporal Transformer Model for Traffic Flow Forecasting

Predictive analytics for traffic flow optimization in urban logistics: A transformer-based time series approach