Classification of synchronous fluorescence of petroleum oils

摘要

A pattern classification system for the identification of UV-visible synchronous fluorescence of petroleum oils is developed. The system is a composite of three phases, namely, feature extraction, feature selection and pattern classification. These phases are briefly described, focusing particularly on the classification method. A method called successive feature elimination process (SFEP) is used for feature selection and a proximity index classifier (PIC) is developed for classification. The feature selection method extracts as many features from spectra as conveniently possible and then applies the SFEP process to remove the redundant features. From the remaining features a significantly smaller feature subset is selected that enhances the recognition performance of the PIC classifier. The SFEP and PIC methods are formally described. These methods are successfully applied to the classification of UV-visible synchronous fluorescence spectra. The features selected by the algorithm are used to classify twenty different sets of petroleum oils. The system was trained on the design set on which the recognition performance was 100%. The performance on the testing set was over 93% by successfully identifying 28 out of 30 samples in six classes. This performance is very encouraging. In addition, the method is computationally inexpensive and is equally useful for large data set problems as it always partitions the problem into a set of two class problems.