ABSTRACT
The theory of probability is the oldest and best established method for modelling uncertainty and hence is the starting point of the discussion of uncertainty. The description of the theory of belief functions assumes some familiarity with probability theory. This chapter reviews the basics of probability theory, especially emphasizing Bayesian ways of thinking about probability. It provides a rapid review of the concepts of probability from the Bayesian viewpoint, paying particular attention to the central role of conditional probability. Conditional probability statements play the same role in graphical models that logical rules play in a rule-based expert system and a good understanding of conditional probability is a prerequisite for understanding many of the models. The chapter describes two simple probability models, the Bernoulli and Poisson processes, which are the building blocks of the reliability models. It shows how probability is used in Bayesian modelling of common statistical problems.
