Minimisation of decision errors in a probabilistic neural network for change point detection in mechanical systems

Probabilistic neural nets have been applied in the detection of structural damage. These networks, which rely upon approximating the multivariate dens ity of the training data, have been shown to be effective in some applicati ons. However, quantification of decision errors, which must ultimately be u sed to rank their effectiveness, has received little attention. In this pap er, a two-dimensional probabilistic pattern classifier (PPC) based on an L- 2 detector is studied. Each dimension is modelled as a Gaussian random vari able in order to directly study the effects introduced by a commonly used d ensity estimator. The region of acceptance is examined for different parame ters of the kernel density estimator. The detector behaviour of the origina l PPC is compared to theory. It is demonstrated that performance is affecte d by errors introduced by a kernel constructed using a finite set of data, and also from theoretical limitations inherent in the test statistic. A mod ification of the test statistic is suggested to improve the sensitivity for structural damage detection. (C) 1999 Academic Press.