In this thesis, we study sequential decision-making under uncertainty in the context of smart grids using reinforcement learning. The underlying mathematical model for reinforcement learning algorithms are Markov Decision ...