Empirical modeling methods using partial data

Methods were developed to calculate empirical models for device error behavior from data sets with missing data. These models can be used to develop reduced test point testing procedures for the devices. Normally, models are built from only full data measurement sets, and partial data sets are discarded. For models built from noisy data, the accuracy of the models improves as more data is used. This paper explores methods to use partial data sets. Both real and simulated data results are described. Simulations show that the proposed partial data methods improve the accuracy of the models for some test points. When these methods are applied to real data where the underlying model has changed, the improvement is less than the simulations predict.