Project

In my project, I will be extending reinforcement learning to encorporate higher order learning. Reinforcement learning is a machine learning paradigm where an agent interacts with its environment and modifies its actions to maximize its reward. Higher Order Learning is a classification mechanism that connects relationships between attributes found in different datasets of the same class. This data can come in many forms. Our approach is to see if higher order learning can be used to classify states found in RL. We will test this method on the traditional domains used in RL, as well as some non-traditional RL domains, such as nuclear detection.