A stable and effective learning strategy for trainable greedy decoding

Mathematics

• Learning Strategies 100%
• Beam Search 88%
• Decoding 65%
• Actors 41%
• Corpus 40%
• Computational Cost 30%
• Neural Networks 28%
• Machine Translation 26%
• Reinforcement Learning 22%
• Search Strategy 20%
• Architecture 16%
• Target 12%
• Experiment 11%
• Output 11%
• Metric 9%
• Model 9%
• Range of data 9%

Engineering & Materials Science

• Decoding 50%
• Neural networks 36%
• Reinforcement learning 30%
• Costs 22%
• Experiments 10%