Decision Making in the Presence of Measurement Uncertainty: an Approach in Terms of the Theory of Evidence

In almost every scientific and technical activity, decisions are taken on the basis of data coming from the field, either in a direct way (measurement results) or in an indirect way (as the result of some processing of the experimental data). Generally, decisions are taken by comparing data coming from the field with each other or with a given threshold. The problem is that experimental data, collected as the results of a measurement activity, are affected by uncertainty, and therefore the comparison cannot be performed as a simple comparison between two scalar values, but should also take into account the measurement uncertainty. In fact, neglecting measurement uncertainty could lead to decisions that might be dramatically wrong. The problem of decision making in the presence of measurement uncertainty is very often neglected, or approached in a probabilistic way. This paper proposes a new, more general approach, based on the theory of evidence