Scholarly record

THE NEUTRALIZATION OF SYNGENETIC COMPONENT EFFECT IN STREAM SEDIMENTS GEOCHEMICAL EXPLORATION USING ARTIFICIAL NEURAL NETWORKS (CASE STUDY: SHIRINKAND 1:50000 SHEET)

M. Ahadi, A.HEZARKHANI

Abstract

Variability of stream sediment survey data have two principal component including syngenetic component relates to lithogenesis and lithogeochemical variation, and epigenetic component relates to mineralization which its determination is the target in geochemical exploration. In regional stream sediment surveys, effective lithologic units in syngenetic component are located in the upstream of related sample. Various methods are used for neutralization of syngenetic component in stream sediment geochemical surveys such as lithologic groups separation, principal component analysis, fuzzy logic and artificial neural network. This paper describes application of comparative neural network in naturalization of syngenetic component which is very suitable approach for data clustering. After data clustering by this method based on rock forming elements and validation of results, 8 optimum clusters were obtained that indicated lithologic units of survey area, corresponding on its related geology map. After data normalization of each cluster to its center, enrichment index data file were compiled, then unielement anomalies were indicated using computation of average plus standard deviation for enrichment index data that highlighted abnormal areas.

Publication details

Back to publication list