In many situations we have to make decisions with a lack of knowledge. This is the case in e.g. traditional military intelligence, and cyber threat intelligence where there are established methods to reason under uncertainty. This is also the case with AI and ML where autonomous systems make decisions under uncertainty and often with time constraints.
ML literature distinguishes between aleatoric uncertainty resulting from inherent stochastic dependencies, and epistemic uncertainty resulting from lack of knowledge. Traditional logics assume discrete input arguments without uncertainty, while probabilistic reasoning adds the dimension of aleatoric uncertainty. Subjective logic handles epistemic uncertainty in addition, as well as trust and subjective beliefs of different agents.
This seminar presents the theory of subjective logic and applications of reasoning under uncertainty with representations as Beta and Dirichlet distributions.