INTEGRATING MULTIVARIATE STATISTICAL-ANAL YSIS WITH NEURAL NETWORKS FOR PATTERN-CLASSIFICATION OF COMPLEX CHEMICAL INFORMATION

A new method by integrating the multivariate statistical analysis with neural network used for complex pattern classification was proposed i n this paper, First, a particularly developed statistical method. call ed correlational components analysis was employed to extract pattern c haracteristics from the original sample pattern space, These pattern c haracteristics were then used as inputs to a multi-layered feedforward neural networks for further pattern classification. The proposed appr oach transforms the complex patterns into lower dimensional and mutual ly decoupled ones, it also takes the advantages of the self-learning c apability of the neural networks. Finally, a practical example of natu ral spearmint oil was used, to verify the effectiveness of the new met hod. The results showed that the proposed integrated approach gives be tter results than other conventional methods.