GCGAN: Generative Adversarial Nets with Graph CNN for Network-Scale Traffic Prediction

Keyphrases

• Adversarial Training 66%
• Arterial Road Network 33%
• Chicago 33%
• Deep Learning Methods 33%
• Deep Neural Network 33%
• Deep Neural Network Algorithm 33%
• Encoder-decoder 33%
• Feature Learning 33%
• Generative Adversarial Nets 100%
• Graph CNN 100%
• Graph Convolutional Network 33%
• Network Scale 100%
• Neural Network Method 33%
• Prediction Framework 33%
• Real Application 33%
• Real-time Traffic 33%
• Road Network 66%
• Road Segment 100%
• Seq2Seq 33%
• Sequence-to-sequence 33%
• Spatial Correlation 33%
• Statistical Methods 66%
• Time-varying Characteristics 33%
• Traditional Statistics 33%
• Traffic Condition 66%
• Traffic Congestion 33%
• Traffic Data 33%
• Traffic Dataset 33%
• Traffic Prediction 100%
• Traffic Prediction Model 66%
• Training Loss 33%
• Transportation Network 66%
• Urban Areas 33%
• Urban Transportation System 33%

Computer Science

• Adversarial Machine Learning 40%
• Based Prediction Framework 20%
• Convolutional Neural Network 100%
• Deep Learning Method 20%
• Deep Neural Network 40%
• Discriminator 20%
• Generative Adversarial Networks 100%
• Graph Convolutional Network 20%
• Prediction Model 40%
• Real Application 20%
• Representation Learning 20%
• Seq2Seq 40%
• Spatial Correlation 20%
• Traffic Condition 40%
• Traffic Prediction 100%
• Transportation Network 40%
• Urban Transportation System 20%