Event Source Page Discovery via Policy-Based RL with Multi-task Neural Sequence Model

Computer Science

• Training Phase 100%
• Reinforcement Learning 100%
• Unlabeled Data 100%
• Deep Reinforcement Learning 100%
• Sequence Model 100%
• Tuning Phase 100%
• Limited Labeled Data 100%

Keyphrases

• Reinforcement Learning 100%
• Event Sourcing 100%
• Neural Sequences 100%
• Systems-based 20%
• Training Phase 20%
• Reward System 20%
• Policy-based 20%
• Unlabeled Data 20%
• Unit Cost 20%
• Adaptive Threshold 20%
• Number of Steps 20%
• Deep Reinforcement Learning Model 20%
• Limited Labeled Data 20%
• Real-word Data 20%