Optimized Graph Convolution Recurrent Neural Network for Traffic Prediction

Guo, Kan; Hu, Yongli; Qian, Sean; Líu, Hao; Zhang, Ke; Sun, Yeneng; Gao, Junbin; Yin, Baocai

doi:10.1109/tits.2019.2963722

Cited by 259 publications

(117 citation statements)

References 34 publications

Supporting

Mentioning

100

Contrasting

Order By: Relevance

“…For this reason, an efficient adaptive demodulator was suggested, combined with DL. So, an improved interleaved multi-coding technique is exploited, and an advanced mapping scheme is proposed to provide better communication performance under different AT conditions with DL methods [11]. In this section, we employ adaptive demodulation using two DL techniques to classify and predict distinct videos using different datasets of encrypted multi-coded N -OAM-SK states.…”

Section: The Adaptive Detection Mechanism Of Oam-sk Using Two DL Methodsmentioning

confidence: 99%

“…This layer slips cubic shape convolution filters to the 3D input and then convolves the input with the filters by moving the filters along the input horizontally, vertically, and alongside the depth. The down-sampling process is performed using a 3D max pooling layer by splitting three-dimensional input into cuboidal pooling sections and computing the maximum of each region [11], [12].…”

Section: B 3d-cnn Modelmentioning

confidence: 99%

“…Recently, the absenteeism of twirls in convolution neural networks (CNNs), and feed forward neural networks (FFNNs), has made them insufficient techniques for use in sequence labeling [11]. Recurrent neural networks (RNNs) have been used for superior exploration of the temporal information of sequential data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Performance Analysis of 3D Video Transmission Over Deep-Learning-Based Multi-Coded N-ary Orbital Angular Momentum FSO System

et al. 2021

View full text Add to dashboard Cite

Orbital angular momentum-shift keying (OAM-SK), which is the rapid switching of OAM modes, is vital but seriously impeded by the deficiency of OAM demodulation techniques, particularly when videos are transferred over the system. Thus, in this paper, 3D chaotic interleaved multi-coded video frames (VFs) are conveyed via an N-OAM-SK free space optic (FSO) communication system to enhance the reliability and efficiency of the model. To tackle the defects of the OAM-SK-FSO mechanism, two efficient deep learning (DL) techniques, namely convolution recurrent neural network (CRNN) and 3D convolution neural network (3DCNN) are used to decode OAM modes with a lower BER. Moreover, a graphics processing unit (GPU) is used to accelerate the training process with slight power consumption. The utilized datasets for OAM states are generated by applying different scenarios using trial and error method. The simulation results imply that LDPC coded VFs achieve the greatest peak signal-to-noise ratios (PSNRs) and the lowest BERs using the 16-OAM-SK model. Both 3DCNN and CRNN techniques have nearly the same performance, but this performance deteriorates in the case of larger dataset classes. Moreover, the GPU accelerates the performance by almost 67.6% and 36.9% using CRNN and 3DCNN techniques, respectively. These two DL techniques are more effective in evaluating the classification accuracy than the other traditional techniques by almost 10 − 20%.

show abstract

Section: The Adaptive Detection Mechanism Of Oam-sk Using Two DL Methodsmentioning

confidence: 99%

Section: B 3d-cnn Modelmentioning

confidence: 99%

See 1 more Smart Citation

Performance Analysis of 3D Video Transmission Over Deep-Learning-Based Multi-Coded N-ary Orbital Angular Momentum FSO System

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Current lines of research exploiting graphs in the transportation domain are mainly focused on static topological information related to nodes and intersections [8][9][10], so neglecting the dynamic information coming from sensors. Among the few exceptions in that sense are solutions based on graph convolution neural-network based approaches (e.g., [7,11]), which however are still at very early stages of development and suffer from extremely high computation times, thus being rather inappropriate for large-scale real-time traffic monitoring [12]. Among the metrics for complex networks that take into account both network topology and traffic dynamics, we propose dynamic Betweenness Centrality (BC), i.e., BC continuously computed over a dynamic graph.…”

Section: Introductionmentioning

confidence: 99%

Graph-based ahead monitoring of vulnerabilities in large dynamic transportation networks

et al. 2021

View full text Add to dashboard Cite

Betweenness Centrality (BC) has proven to be a fundamental metric in many domains to identify the components (nodes) of a system modelled as a graph that are mostly traversed by information flows thus being critical to the proper functioning of the system itself. In the transportation domain, the metric has been mainly adopted to discover topological bottlenecks of the physical infrastructure composed of roads or railways. The adoption of this metric to study the evolution of transportation networks that take into account also the dynamic conditions of traffic is in its infancy mainly due to the high computation time needed to compute BC in large dynamic graphs. This paper explores the adoption of dynamic BC, i.e., BC computed on dynamic large-scale graphs, modeling road networks and the related vehicular traffic, and proposes the adoption of a fast algorithm for ahead monitoring of transportation networks by computing approximated BC values under time constraints. The experimental analysis proves that, with a bounded and tolerable approximation, the algorithm computes BC on very large dynamically weighted graphs in a significantly shorter time if compared with exact computation. Moreover, since the proposed algorithm can be tuned for an ideal trade-off between performance and accuracy, our solution paves the way to quasi real-time monitoring of highly dynamic networks providing anticipated information about possible congested or vulnerable areas. Such knowledge can be exploited by travel assistance services or intelligent traffic control systems to perform informed re-routing and therefore enhance network resilience in smart cities.

show abstract

“…Convolution neural network (CNN) has achieved great success in processing data on Euclidean domains (grid structure) [1,2], such as image, speech, and video [3][4][5][6]. Therefore, many recent studies have devoted to the extension of convolution operations to the data on non-Euclidean domains and proposed the graph convolution network (GCN) [7][8][9].Then GCN has been successfully used to achieve improvements in several research fields, such as protein interface [10], action recognition [11], and traffic data processing [12]. The GCN mainly includes the spectral-based and spatial-based methods, which can leverage the topology information of the graph data to aggregate the local node features, and then automatically learn the embeddings for graph nodes.…”

Section: Introductionmentioning

confidence: 99%

Uniform Pooling for Graph Networks

et al. 2020

View full text Add to dashboard Cite

The graph convolution network has received a lot of attention because it extends the convolution to non-Euclidean domains. However, the graph pooling method is still less concerned, which can learn coarse graph embedding to facilitate graph classification. Previous pooling methods were based on assigning a score to each node and then pooling only the highest-scoring nodes, which might throw away whole neighbourhoods of nodes and therefore information. Here, we proposed a novel pooling method UGPool with a new point-of-view on selecting nodes. UGPool learns node scores based on node features and uniformly pools neighboring nodes instead of top nodes in the score-space, resulting in a uniformly coarsened graph. In multiple graph classification tasks, including the protein graphs, the biological graphs and the brain connectivity graphs, we demonstrated that UGPool outperforms other graph pooling methods while maintaining high efficiency. Moreover, we also show that UGPool can be integrated with multiple graph convolution networks to effectively improve performance compared to no pooling.

show abstract

Optimized Graph Convolution Recurrent Neural Network for Traffic Prediction

Cited by 259 publications

References 34 publications

Performance Analysis of 3D Video Transmission Over Deep-Learning-Based Multi-Coded N-ary Orbital Angular Momentum FSO System

Performance Analysis of 3D Video Transmission Over Deep-Learning-Based Multi-Coded N-ary Orbital Angular Momentum FSO System

Graph-based ahead monitoring of vulnerabilities in large dynamic transportation networks

Uniform Pooling for Graph Networks

Contact Info

Product

Resources

About