Dynamic Sparse Training for Deep Reinforcement Learning

Computer Science

• Deep Reinforcement Learning 100%
• Networks 80%
• Learning Agent 40%
• Decision-Making 20%
• Topology 20%
• Learning Algorithm 20%
• Computation 20%
• Algorithms 20%
• Good Performance 20%
• Learning Process 20%
• Experimental Result 20%
• Speed-up 20%
• Reinforcement Learning 20%
• Computation Cost 20%
• Continuous Control 20%
• Training Process 20%
• Floating-Point Operation 20%
• Incoming Data 20%