Deeply-supervised nets

Mathematics

• Transparency 46%
• Training 34%
• Objective function 26%
• Learning Process 22%
• Stochastic Gradient 21%
• Stochastic Methods 18%
• Gradient Method 18%
• Architecture 17%
• Experimental Results 16%
• Neural Networks 15%
• Robustness 13%
• Minimise 12%
• Gradient 12%
• Output 11%
• Strategy 11%
• Distinct 11%
• Performance 9%

Engineering & Materials Science

• Transparency 100%
• Gradient methods 58%
• Convolutional neural networks 43%