Identification of the 'Ambiguity Features' of the Conodont by the Artificial Neural Network
Hua Chai · Dizhi ke-ji qingbao · 2009
Conodont is an important kind of fossil in the stratigraphy research.But ever since a long time ago the identification of the conodont has depended on the experience of the paleontologists because many of their features are significant but hard to be identified quantitatively.This paper describes these features with non-linear method in mathematics and designs a set of multiple hidden layer feed forward network based on the back propagation algorithm to identify those 'ambiguity features'.The method has yielded good effect,and laid a foundation for the further numerical research on the conodont.