To map characteristic space onto mapping space(or identification space), an impor tant problem is to select mappingpoints in mapping space for the multianomaly identification method.

The geochemical research of soil hydrocarbon geochemical test profile in Ejina Qi region shows that the geochemical indices of acid hydrocarbon display higher hydrocarbon background value and stronger hydrocarbon anomaly,being characterized by heavier methane carbon isotope and bigger drying coefficient.

This paper presents an entropy-based large-scale network traffic anomaly classification method for the integrated use of the subspace method and the k-means clustering method.

Based on the characterization of anomalies and interference frequency in geochemical exploration, a new method to eliminate interference by using wavelet analysis is established, and a theoretical model of anomaly types and their formation mechanisms is developed.