Simple random search provides a competitive approach to reinforcement learning

[obastani]

https://arxiv.org/abs/1803.07055