Chemometric regression: plutonium isotopic composition

^238-241Pu isotopic abundances to predict the ²⁴²Pu abundance using Multiple Linear Regression (MLR), Principle Component Regression (PCR) and Partial Least Squares Regression (PLSR) algorithms. The MLR algorithm was found to be the best. The effect of including ²³⁸Pu information on the ²⁴²Pu prediction is small but significant especially for the high accuracy. Comparison of the MLR results with those obtained by employing five reported empirical methods reveals superior prediction capability of MLR models. Among these empirical models, the best prediction capability was found for Bignan correlation, interestingly the only model which uses ²³⁸Pu abundance data for ²⁴²Pu correlation. The results clearly demonstrate the fact that the production routes of ²³⁸Pu and ²⁴²Pu have a definite correlation and the use of ²³⁸Pu abundance data in ²⁴²Pu prediction is important to get accurate results