Simple Random Search Provides a Competitive Approach to Reinforcement Learning Ouch, Ben Recht at it again. I won't be writing detailed notes as the paper is easy to read. http://www.argmin.net/2018/03/20/mujocoloco/