An Interpretable LSTM Network for Solar Flare Prediction

Keyphrases

• Active Regions 100%
• Solar Flare Prediction 100%
• Long Short-term Memory Network 100%
• Prediction Model 66%
• Model Prediction 66%
• Memory-based 33%
• Temporal Dynamics 33%
• Modeling Data 33%
• National Center 33%
• Environmental Information 33%
• Deep Learning Model 33%
• Flare Prediction 33%
• Attention Mechanism 33%
• Magnetic Parameters 33%
• Behavioral Model 33%
• M-class 33%
• X-ray Flares 33%
• Input Sequence 33%
• X-class Flare 33%
• Geostationary Operational Environmental Satellite 33%
• Multiple Sequences 33%
• Long Short-term Memory 33%
• Lack of Transparency 33%
• Model Explanation 33%
• Black-box Model 33%

Computer Science

• Long Short-Term Memory Networks 100%
• Prediction Model 66%
• Model Prediction 66%
• Environmental Information 33%
• Interpretability 33%
• Attention (Machine Learning) 33%
• Behavior Model 33%
• Deep Learning Model 33%
• Contributing Factor 33%
• Temporal Dynamic 33%
• Long Short-Term Memory Network 33%
• Predicted Output 33%

Engineering

• Long Short-Term Memory 100%
• Active Region 75%
• Data Sample 75%
• Model Prediction 50%
• Deep Learning 25%
• Input Sequence 25%
• Model Data 25%
• Black-Box Model 25%
• Internal Working 25%