Rare Fault Detection by Possibilistic Reasoning

Abstract

Kernel based neural networks with probabilistic reasoning are suitable for many practical applications. But influence of data set sizes let the probabilistic approach fail in case of small data amounts. Possibilistic reasoning avoids this drawback because it is independent of class size. The fundamentals of possibilistic reasoning are derived from a probability/possibility consistency principle that gives regard to relations. It is demonstrated that the concept of possibilistic reasoning is advantageous for the problem of rare fault detection, which is a property desired for semiconductor manufacturing quality control.