Improving materials property predictions for graph neural networks with minimal feature engineering

Engineering

• Engineering 100%
• Band Gap 66%
• Nodes 33%
• Functional Property 33%
• Learning Approach 33%
• Application Domain 33%
• Formation Energy 33%
• Gaussians 33%
• Network Model 33%
• Plane Wave 33%
• Learning System 33%

Computer Science

• Graph Neural Network 100%
• Neural Network Model 33%
• Superior Performance 33%
• Machine Learning Approach 33%
• Attention (Machine Learning) 33%
• Research Material 33%
• Plane Function 33%
• Predictive Power 33%
• Application Domain 33%
• Functional Property 33%