Investigating Recurrence and Eligibility Traces in Deep Q-Networks
This work addresses improving training efficiency and performance in reinforcement learning for Atari games, but it appears incremental as it combines existing techniques.
The paper investigated combining eligibility traces with recurrent networks in Deep Q-Networks for Atari games, showing benefits in some games and highlighting the importance of optimization in training.
Eligibility traces in reinforcement learning are used as a bias-variance trade-off and can often speed up training time by propagating knowledge back over time-steps in a single update. We investigate the use of eligibility traces in combination with recurrent networks in the Atari domain. We illustrate the benefits of both recurrent nets and eligibility traces in some Atari games, and highlight also the importance of the optimization used in the training.