On the convergence of a class of Adam-type algorithms for non-convex optimization

Keyphrases

• Adagrad 100%
• Adams Method 100%
• Adaptive Gradient 50%
• AMSGrad 50%
• Convergence Analysis 50%
• Convergence Behavior 50%
• Convergence Rate 50%
• First-order 50%
• Gradient-based 50%
• Learning Rate 50%
• Momentum Algorithm 50%
• Nonconvex 50%
• Nonconvex Optimization 100%
• Nonconvex Problem 50%
• Nonconvex Stochastic Optimization 50%
• Popular 50%
• Search Direction 50%
• Training Deep Neural Networks 50%

Computer Science

• Analysis Framework 100%
• Convergence Behavior 100%
• Convergence Rate 100%
• Convex Optimization 100%
• Deep Neural Network 100%
• Learning Rate 100%
• Mathematical Convergence 100%
• Search Direction 100%
• Stochastic Optimization 100%
• Sufficient Condition 100%

Mathematics

• Convergence Analysis 50%
• Convergence Behavior 50%
• Convergence Rate 50%
• Deep Neural Network 50%
• Nonconvex Problem 50%
• Open Question 50%
• Search Direction 50%
• Stochastics 50%
• Sufficient Condition 50%
• Type Method 100%