Problem 23. Deep Reasoning Reinforcement Learning for Cognitive Information Warfare

Cognitive Information warfare has become a paradigm that intelligence organisations need toincreasingly take into consideration as information become pervasive and abundant in every aspect ofsociety. Reinforcement learning (RL) is emerging as a powerful approach in processing large amountsof information for decision-making but still has limitations in terms of requiring a mammoth amount ofcomputation, data and time to achieve results. This is due in large part to the fact that RL often employsa trial and error approach starting from scratch without any prior knowledge built in. Yet in manydomains, there is human knowledge and expertise available that could be used to guide (i.e. warmstart)the RL algorithms to help increase efficiency.