A partially observable multi-ship collision avoidance decision-making model based on deep reinforcement learning

Earth and Planetary Sciences

• Model 100%
• Learning 100%
• Ship 100%
• Decision Making 100%
• Algorithms 50%
• Policy 50%
• Optimization 50%
• Complexity 33%
• Decision 33%
• Efficiency 33%
• Research 16%
• State 16%
• Foundation 16%
• Show 16%
• Simulation 16%
• Fitting 16%
• Observation 16%
• Attention 16%
• Accuracy 16%
• Information 16%
• Guideline 16%
• Safety 16%
• Error 16%
• Route 16%

Computer Science

• Decision-Making 100%
• Models 100%
• Collisions 100%
• Deep Reinforcement Learning 100%
• Complexity 40%
• Markov Decision Process 40%
• Optimization Policy 40%
• Simulation 20%
• Algorithms 20%
• Accuracy 20%
• Functions 20%
• Mathematical Optimization 20%
• Conventional Algorithm 20%
• Training Model 20%