Possibility and necessity pattern classification using an interval arithmetic perceptron

In the work presented in this paper, an Interval Arithmetic Perceptron (IAP ) is used to detect the region in the input space to which an uncertainty d ecision should be appropriately associated. This region may be originated b oth by sub-regions which are not represented in the training set, and by su bregions where the probabilities of the two classes are very similar. To tr ain the IAP, an algorithm will be presented which in particular is abbe det ect the two certainty regions and the uncertainty one. From the interval we ights thus obtained, a confidence interval of the probability will also be evaluated. The algorithm has been used for studying a simple artificial pro blem and two real-world applications, the Iris and Breast Cancer databases. Regarding the latter application in particular a statistical analysis of t he results is presented, together with a discussion of the possible alterna tive classifications of the patterns attributed to the uncertainty region.