Timezone: »

Cold-Start Reinforcement Learning with Softmax Policy Gradient
Nan Ding · Radu Soricut

Tue Dec 05 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #109 #None

Policy-gradient approaches to reinforcement learning have two common and undesirable overhead procedures, namely warm-start training and sample variance reduction. In this paper, we describe a reinforcement learning method based on a softmax value function that requires neither of these procedures. Our method combines the advantages of policy-gradient methods with the efficiency and simplicity of maximum-likelihood approaches. We apply this new cold-start reinforcement learning method in training sequence generation models for structured output prediction problems. Empirical evidence validates this method on automatic summarization and image captioning tasks.

Author Information

Nan Ding (Google)
Radu Soricut (Google)

More from the Same Authors