Defective wafer detection using neural network approach and multivariate statistical approach

In this paper, we introduce an automatic classification approach based on combining two algorithms. One of them belongs to Neural Networks approach and the second is based on Multivariate Statistical approach. In this study we will present two algorithms their relative strengths and weaknesses. We realize that these two methods can complement one another resulting in better decision support system. Integrating these complementary features is one way to develop hybrid system that could overcome the limitations of individual solution strategies. We evaluate this hybrid system on data provided by semiconductor fab, especially on Parametric Tests (PT) data. This procedure was successfully validated at PT data provided by STMicroelectronics - Rousset fab.