Xueguang Lyu

On Improving Decentralized Hysteretic Deep Reinforcement Learning.

2018 · arXiv (Cornell University)

Recent successes of value-based multi-agent deep reinforcement learning employ optimism in value function by carefully controlling learning rate(Omidshafiei et al., 2017) or reducing update prob-ability (Palmer et al., 2018). We introduce a de-centralized quantile estimator: …