Escaping from Zero Gradient: Revisiting Action-Constrained Reinforcement Learning via Frank-Wolfe Policy Optimization

Mathematics

• Reinforcement Learning 100%
• Policy 65%
• Gradient 50%
• Optimization 50%
• Zero 33%
• Update 32%
• Resource Constraints 25%
• Projection 24%
• Gradient Projection 23%
• Training Samples 22%
• Real-world Applications 20%
• Robot 20%
• Learning Algorithm 19%
• Scheduling 16%
• Kinematics 16%
• Benchmark 16%
• Regression 12%
• Experiment 12%
• Demonstrate 9%
• Generalise 9%

Engineering & Materials Science

• Reinforcement learning 66%
• Learning algorithms 18%
• Kinematics 17%
• Scheduling 15%
• Robots 14%
• Experiments 7%