Patent US 7,107,048

In lieu of the common practice of attempting to maximize the likelihood of sample deviations from unknown true values, representations of data inversions are generated by implementing optimizing techniques which are based upon a concept of maximizing the likelihood of deviations of successively determined inversion-conforming data sets from previously acquired (hence invarient) data- point samples. Inversion-conforming data sets are considered to be sets of coordinate points which correspond to the projection of respective data points along corresponding coordinates onto a fitting approximation or onto the locus of any respectively rendered approximating relationship which corresponds to a data inversion or a considered estimate of the same.