Search

The analysis of dietary fiber is done by gravimetric method where the fiber is separated and weighed, and by component analysis process during which dietary fiber components are estimated and integrated. For separation of soluble and insoluble fibres the enzymic, gravimetric and component analysis are performed. An important component of dietary fiber is resistant starch (RS). The three forms of RS are physically enclosed starch, raw starch granules of the B-type and retrograded amylose.

An examination on the role of artificial neural networks (ANN) for milk shelf-life revealed that that were useful for predicting milk shelf-life when applied to gas chromatographic profiles with sensory-based shelf-life. The predictive ability were better than the principal components regression. It is a more amenable for quality control situations. The principle components in this network are calculated before regression analysis is applied.

A neural network model, APEX, associated with multiple principal component extraction, provided solutions for several principal component analysis applications. The normalized Hebbian learning rule was used to train the synaptic weights of the model which comprised output units having a hierarchical set of lateral connections. Such neurocomputation modeling provided information regarding biological perceptual processing.