Whenever you're ready, walk me through how you'd approach this.
What is batch normalization, and how would you optimize a reinforcement learning algorithm?
