Do Pre-processing and Augmentation Help Explainability? A Multi-seed Analysis for Brain Age Estimation

Mathematics

• Preprocessing 100%
• Brain 94%
• Training 73%
• Augmentation 61%
• Overfitting 46%
• Registration 45%
• Performance 41%
• Performance Model 25%
• Data Augmentation 23%
• Learning Algorithm 20%
• Black Box 19%
• Availability 18%
• Learning 15%
• Efficient Algorithms 14%
• Output 12%
• Context 10%
• Generalization 8%
• Model 5%

Engineering & Materials Science

• Seed 89%
• Brain 71%
• Deep learning 30%
• Learning algorithms 14%
• Availability 12%