SYSTEM AND METHOD FOR NON-PARAMETRIC MODELING OF PROCESSED INDUCED VARIABILITY

Method for adequately modeling process induced variabilities is disclosed that comprises the steps of acquiring experimental data and defining a particular design space. Values for the mean and standard deviation of the experimental data at each of the points defining the design space are calculated. The experimental values of the output parameters at each of the design points is normalized to extract the shape of the distribution of each of the design points. The normalized values are then merged to form a cumulative distribution function associated with the data. The cumulative distribution function is applied to a new design point in a predicted fashion by first calculating a mean and standard deviation value for the new point by interpolating from the mean and standard deviation values from the experimental data. The cumulative distribution function is then scaled and centered using the interpolated mean and standard deviation values to provide a predicted data distribution for the new design point.