Using PCA to Efficiently Represent State Spaces

William Curran, Tim Brys, Matthew Taylor +1

Reinforcement learning algorithms need to deal with the exponential growth of states and actions when exploring optimal control in high-dimensional spaces. This is known as the cur…