A unified temporal link prediction framework based on nonnegative matrix factorization and graph regularization

Keyphrases

• Non-negative Matrix Factorization 100%
• Dynamic Networks 100%
• Prediction Framework 100%
• Graph Regularization 100%
• Temporal Link Prediction 100%
• Matrix Diagram 100%
• Root Mean Square Error 33%
• Network Structure 33%
• Evaluation Metrics 33%
• Temporal Features 33%
• Spatial Features 33%
• Receiver Operating Characteristic Curve 33%
• Factorization Method 33%
• Real-world Systems 33%
• Community Data 33%
• Multiplicative Updating Rules 33%
• Characteristic Roots 33%
• Latent Feature 33%
• Topological Structure 33%
• Modularity Matrix 33%
• Weighted Modularity 33%
• Synthetic Networks 33%
• Weight Matrix 33%
• Historical Network Analysis 33%
• Structure Data 33%

Computer Science

• nonnegative matrix factorization 100%
• Dynamic Network 100%
• Link Prediction 100%
• Regularization 100%
• Prediction Framework 100%
• Evaluation Metric 33%
• Temporal Feature 33%
• Baseline Method 33%
• Characteristic Curve 33%
• Network Structure 33%
• Topological Structure 33%
• Characteristic Root 33%
• Root Mean Squared Error 33%