GCN-GAN: Integrating Graph Convolutional Network and Generative Adversarial Network for Traffic Flow Prediction

Zheng, Hongling; Li, Xiang; Li, Yongfeng; Yan, Ziqin; Li, Tinghong

doi:10.1109/access.2022.3204036

Cited by 18 publications

(9 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The methods of using artificial neural network models to predict short-term traffic flow [6] all have the problem of low prediction accuracy. The urban highway flow prediction model that integrates graph convolutional neural networks and generative adaptive neural networks (GCN-GAN) has weak interpretability [7]. The traffic flow prediction model that integrates LSTM and GCN can only achieve short-term prediction [8].…”

Section: Figure 1 Statistics Of Motor Vehicle Ownership In Chinamentioning

confidence: 99%

An Integrated Convolutional Neural Network-Bidirectional Long Short-Term Memory-Attention Mechanism Model for Enhanced Highway Traffic Flow Prediction

Kan,

Li,

Wang

2024

JUDM

View full text Add to dashboard Cite

The burgeoning expansion of the Internet of Things (IoT) technology has propelled Intelligent Traffic Systems (ITS) to the forefront of IoT applications, with accurate highway traffic flow prediction models playing a pivotal role in their development. Such models are essential for mitigating highway traffic congestion, reducing accident rates, and informing city planning and traffic management strategies. Given the inherent periodicity, nonlinearity, and variability of highway traffic data, an innovative model leveraging a Convolutional Neural Network (CNN), Bidirectional Long Short-Term Memory (BiLSTM), and Attention Mechanism (AM) is proposed. In this model, feature extraction is accomplished via the CNN, which subsequently feeds into the BiLSTM for processing temporal dependencies. The integration of an AM enhances the model by weighting and fusing the BiLSTM outputs, thereby refining the prediction accuracy. Through a series of experiments and the application of diverse evaluation metrics, it is demonstrated that the proposed CNN-BiLSTM-AM model surpasses existing models in prediction accuracy and explainability. This advancement positions the model as a significant contribution to the field, offering a robust and insightful tool for highway traffic flow prediction.

show abstract

Section: Figure 1 Statistics Of Motor Vehicle Ownership In Chinamentioning

confidence: 99%

An Integrated Convolutional Neural Network-Bidirectional Long Short-Term Memory-Attention Mechanism Model for Enhanced Highway Traffic Flow Prediction

Kan,

Li,

Wang

2024

JUDM

View full text Add to dashboard Cite

show abstract

“…Wang et al [34] suggested a model that utilized GRU to produce aggregated spatial-temporal representations. Several models have employed multiple layers of RNN [16], [21], [33], [35], whereas others have used the attention mechanism [27], [28], [36], [37] to capture the long-term relationship in traffic data.…”

Section: A Temporal Feature Extractionmentioning

confidence: 99%

“…As transportation networks are inherently equipped with graph structures, the GNNs have become the most popular spatial feature extraction method for traffic forecasting. Convolutional GNNs have pioneered GNN-based traffic forecasting research, and have been widely used in concurrent models [15]- [17], [22], [28], [29], [31]- [33], [35], [37], [40], [43], [44], [47], [55]- [60], [62]- [64], [67], [71], [75], [95].…”

Section: B Spatial Feature Extraction With Graph Neural Networkmentioning

confidence: 99%

Performance Evaluation of Building Blocks of Spatial-Temporal Deep Learning Models for Traffic Forecasting

Shin,

Yoon

2023

IEEE Access

View full text Add to dashboard Cite

The traffic forecasting problem is a challenging task that requires spatial-temporal modeling and gathers research interests from various domains. In recent years, spatial-temporal deep learning models have improved the accuracy and scale of traffic forecasting. While hundreds of models have been suggested, they share similar modules, or building blocks, which can be categorized into three temporal feature extraction methods of recurrent neural networks, convolution, and self-attention and two spatial feature extraction methods of convolutional graph neural networks (GNN) and attentional GNN. More importantly, the models have been mostly evaluated for their entire architectures with limited efforts to characterize and understand the performance of each category of building blocks. In this study, we conduct an extensive, multifaceted experiment to understand the influence of building block selection on traffic forecasting accuracy, considering environmental characteristics and dataset distributions. Specifically, we implement six traffic forecasting models using three temporal and two spatial building blocks. When we evaluate the models on four datasets with diverse characteristics, the results show each building block demonstrates distinguishable characteristics depending on study sites, prediction horizons, and traffic categories. The convolution models demonstrate higher overall forecasting performance than other models, whereas self-attention models show competitiveness in less frequent traffic categories, transition states, and the presence of outliers. Based on the results, we also suggest an adaptive model evaluation framework for category-wise predictions of test sets based on the performance of the models on validation sets. The results of this evaluation framework demonstrate improved forecasting accuracy at most by 3.7% without further sophistication in existing model architectures. The results enhance the utility of existing models and suggest guidelines for researchers building traffic forecasting model architectures and for practitioners implementing these state-of-the-art techniques in real-world applications.

show abstract

“…GCN excels at extracting structural features from graphs. Zheng et al [21] used GCN to predict traffic flow with the support of historical traffic flow datasets. This method can accurately estimate the future traffic flow for specific time periods and regions.…”

Section: Introductionmentioning

confidence: 99%

A Temporal Ensembling Based Semi-Supervised Graph Convolutional Network for Power Quality Disturbances Classification

Cai,

Wang,

Jiang

2024

IEEE Access

View full text Add to dashboard Cite

With the integration of multiple energy sources into the power grid makes power quality disturbances (PQDs) more complex. Dealing with power quality problems requires automatic classification of PQDs. This paper proposes a novel semi-supervised Graph Convolutional Network (GCN) framework based on Temporal Ensembling for PQDs classification. Considering both short-term and long-term features of PQDs, a Visibility Graph (VG) based graph theory model was adopted to process PQDs to highlight features. In the proposed semi-supervised framework, Graph Convolutional Network was designed to extract features from massive PQDs and classify PQDs automatically. Due to the fact that GCN belongs to supervised learning, it is necessary to label the data in advance. However, labeling is costly and easily lead to human mistake. Therefore, this article introduces the Temporal Ensembling algorithm which provides pseudo labels to reduce the amount of labeled data and has tolerance to incorrect labels. Simulation results prove that the proposed method is capable of noise resistance, tolerates incorrect labels, and has high classification performance in both single and composite PQDs.

show abstract

GCN-GAN: Integrating Graph Convolutional Network and Generative Adversarial Network for Traffic Flow Prediction

Cited by 18 publications

References 42 publications

An Integrated Convolutional Neural Network-Bidirectional Long Short-Term Memory-Attention Mechanism Model for Enhanced Highway Traffic Flow Prediction

An Integrated Convolutional Neural Network-Bidirectional Long Short-Term Memory-Attention Mechanism Model for Enhanced Highway Traffic Flow Prediction

Performance Evaluation of Building Blocks of Spatial-Temporal Deep Learning Models for Traffic Forecasting

A Temporal Ensembling Based Semi-Supervised Graph Convolutional Network for Power Quality Disturbances Classification

Contact Info

Product

Resources

About