We will discuss the notion of evolution stable strategies.  This will setup the ground to connect game theory and evolution algorithms.  https://drive.google.com/file/d/1LFb8loA876s_c3uWgciLvZl4KEjIOtHL/view?usp=sharing

In the context of reinforcement learning we will show that a specific scheme of Monte Carlo control is monotonic if Q(a, pi) is well estimated by the exploration stage.  https://drive.google.com/file/d/11Aa92Mr3nMF1Gxa5r0kIiHfg-9wn_rkI/view?usp=sharing