Reinforcement learning based on movement primitives for contact tasks

Mathematics

• Reinforcement Learning 100%
• Movement 67%
• Neural Networks 62%
• Contact 56%
• Trajectory 52%
• Robot 49%
• Imitation 38%
• Learning 36%
• Gradient Algorithm 32%
• Learning Algorithm 31%
• Policy 26%
• Controller 25%
• Mm 18%
• Random Search 18%
• Control Algorithm 17%
• Recognition Algorithm 17%
• Tolerance 14%
• Contact Problem 13%
• Necessary 9%
• Performance 7%

Engineering & Materials Science

• Reinforcement learning 66%
• Trajectories 46%
• Neural networks 32%
• Learning algorithms 22%
• Robot learning 19%
• Robots 17%
• Controllers 14%
• Deep neural networks 12%
• Demonstrations 10%