TABLE OF CONTENTS In probabilistic terminology an experiment ? is defined as a process whose outcome is a priori unknown to the analyst. The possible outcomes are all a priori known and classified but which one will occur is unknown at the time the experiment is performed. This definition is consistent with the Bayesian view of probability according to which the outcome of an experiment may be deterministic, but at the moment unknown (e.g. the current number of sons of a friend with which one has lost contact long time ago) as well as stochastic (e.g. the result of a dice toss).
To each experiment ? is associated a sample space ?, which represents the set of all possible outcomes of ?. The sample space can be discrete finite (e.g. for an experiment of a coin or dice toss), countably infinite (e.g. the number of persons crossing the street in a given period of time: in principle, it could be infinite and yet be counted) or continuous (e.g. the value of the dollar currency in the year 3012).
An event E is then a group of possible outcomes of the experiment ?, i.e. a subset of ?. In particular, each possible outcome represents an ( elementary) event itself, being a subset of ?. Further, the null set and the sample space ? can also be considered events.
To each event E is possible to associate its complementary event E
