Deep Reinforcement Learning Patents: An Empirical Survey

Abstract

Deep reinforcement learning is a new type of machine learning resulting from the technical convergence of two more mature machine learning methods, deep learning and reinforcement learning. Deep reinforcement learning is important because it is a scalable method for general intelligence – a machine capable of achieving any definable goal. Since 1995, the United States Patent and Trademark Office has granted over 600 patents with claim terms relating to deep reinforcement learning. Yet, while the literature on software patents is visibly scaling – the literature specifically focused on deep reinforcement learning patents is non-existent. First, this Article discusses technical approaches to deep reinforcement learning; second it contributes the first empirical patent review for deep reinforcement learning technologies, including market modeling, legal claims analysis, and valuation strategies.