Abstract

Taxi demand prediction is one of the key factors in making online taxi hailing services more successful and more popular. Accurate taxi demand prediction can bring various advantages including, but not limited to, enhancing user experience, increasing taxi utilization, and optimizing traffic efficiency. However, the task is challenging because of complex spatial and temporal dependencies of taxi demand. In addition, relationships between non-adjacent regions are also critical for accurate taxi demand prediction, whereas they are largely ignored by existing approaches. To this end, we propose a novel graph and time-series learning model for city-wide taxi demand prediction in this paper. It has two main building blocks, the first one utilize a graph network with attention mechanism to effectively learn spatial dependencies of taxi demand in a broader perspective of the entire city, and the output at each time interval is then transferred to the second block. In the graph network, the edge is defined by an Origin–Destination relation to capture non-adjacent impacts. The second one uses a neural network which is adept with processing sequence data to capture the temporal correlations of city-wide taxi demand. Using a large, real-world dataset and three metrics, we conduct an extensive experimental study and find that our model outperforms state-of-the-art baselines by 9.3% in terms of the root-mean-square error.

Highlights

  • Online taxi hailing services such as Uber, Lyft, and DiDi Chuxing have become an important part of today’s intelligent transportation system (ITS)

  • Starting with graph signal processing (GSP), studies about graph signals gradually divided into two categories: those based on spectral graph theory [37] and those based on graph wavelet theory [38]

  • ST-ResNet still performs worse than Origin–Destination-based Temporal Graph Attention Network (OD-TGAT), which attest the effectiveness of the graph networks in OD-TGAT

Read more

Summary

Introduction

Online taxi hailing services such as Uber, Lyft, and DiDi Chuxing have become an important part of today’s intelligent transportation system (ITS). Model-based approaches assume that taxi demand follows certain predefined patterns. Efforts are made to solely leverage CNNs or its variants [14,15] Despite these achievements, existing deep learning approaches still have the following limitations:. They only consider the spatial influences between adjacent areas, as demonstrated, while the influence between non-adjacent areas is rarely considered The latter spatial influence is crucial for city-wide taxi demand prediction. Different from the previous methods, we model the city as a graph and exploit the influences between non-adjacent areas sufficiently by defining the edges via OD relations in the graph. We propose the Origin–Destination-based Temporal Graph Attention Network (OD-TGAT), in which Graph Attention Networks (GAT) is employed to capture spatial relationships with different edge weights.

Taxi Demand Prediction
Traditional Methods
Deep-Learning-Based Methods
Graph-Convolutional-Based Networks
Region Partition
Time Partition
City Graph
Taxi Demand
Citywide Taxi Demand Prediction
Graph Convolutional Networks
Methodology
Data Description
Data Cleaning
Data Featuring
Framework
Data Transformation
OD-GAT
Temporal Prediction Model
The Experimental Setup
Metrics
Experimental Setting
Comparison with Baseline Models
Performance at Different Time Moments
Parameter Analysis
Findings
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call