A multilayer feed‐forward neural network, trained with a gradient descent, back‐propagation algorithm, is used to estimate the favourability for gold deposits using raster GIS database Tenterfield 1:100 000 sheet area, New South Wales. The consists of solid geology, regional faults, airborne magnetic and gamma‐ray survey data (U, Th, K total count channels), 63 deposit occurrence locations. Input network feature vectors formed by combining values from co‐registered grid cells in each...

The Fe concentrations in fluids present very saline fluid inclusions from the King Island W skarn, hole 16, F-Sn-W deposit and Mary Kathleen U-rare earth element skarn have been determined using phase volume method. Fe-bearing daughter minerals include magnetite, pyrite, amarantite (Fe (super +3) (SO 4 )(OH) . - 3H 2 O), hydrated ferrous chloride. inclusion liquid compositions are inferred a combined analysis of first melting temperatures (= eutectic salt system), crystal identification by...

The effectiveness of some novel software tools used for clustering and classifying multivariate data is tested to evaluate mineral exploration criteria by examining a deposit major fault database. database containing 364 diverse deposits divided into natural groups utilising vector quantisation data-mining approach based on self-organising map (SOM), phenetic cladistic analysis packages. last two approaches are loosely biological principles numerical taxonomy evolutionary relationships,...

Abstract This paper presents a study of neural networks and version spaces for classification remote sensing data. In the first network, precomputed textures based on Spatial Grey Level Dependence (SGLD) method are fed to net in conjunction with spectral The second system is sliding window network which uses all pixels small neighbourhood central pixel. third candidate elimination implementation space concept acquisition shown achieve performance similar that systems but faster training...

Abstract The Hole 16 deposit is a small unexposed F-Sn-W exoskarn with underlying associated endoskarn, greisenized granite and largely ungreisenized Carboniferous 'Elizabeth Creek' granite. Sphalerite geobarometry indicates that this was high level intrusion. skarn formed above cusp mantled by pure marble. Assemblages textures representing successive stages of genesis are: (1) massive andradite, (2) Sn-rich garnet + magnetite ± clinopyroxene fluorite, (3) 'wrigglite' which refers to...

Abstract In the Upper Murray Valley, Victoria, Late Silurian, high‐Si igneous rocks, which are closely associated with alkalic, basaltic dykes, were emplaced at high crustal levels following peak of Benambran Orogeny, deformed and metamorphosed Wagga Zone in Ordovician‐Early Silurian times. These informally termed 'the magmatic suite', include leucogranites, rhyolite dykes flows, ash‐flow tuffs characterised by features. They transitional from mildly peraluminous to metaluminous; they...

This paper describes the integration of neural network ensembles and interval neutrosophic sets using bagging technique for predicting regional-scale potential mineral deposits as well quantifying uncertainty in predictions. Uncertainty types error vagueness are considered this paper. Each component ensemble consists a pair networks trained degrees favourability deposit barren. They truth-membership false-membership values, respectively. Errors occurred prediction estimated multidimensional...

Quantification of uncertainty in mineral prospectivity prediction is an important process to support decision making exploration. Degree uncertainly can identify level quality the prediction. This paper proposes approach predict degrees favourability for gold deposits together with quantification Geographic information systems (GIS) data applied integration ensemble neural networks and interval neutrosophic sets, three different network architectures are used this paper. The its represented...

In the mining industry, identifying new geographic locations that are favorable for mineral exploration is very important. However definitive prediction of such not an easy task. recent years artificial neural networks have received much attention in this area. This paper uses a class known as polynomial network (PNN) to construct model correctly classify given location into deposit and barren areas. information systems (GIS) data location. The method tested on GIS Kalgoorlie region Western...

The paper reports on a pilot study the use of an artificial neural network in conjunction with geographic information system (GIS) for integration large multi-source data sets used regional mineral exploration and prediction prospectivity. A multilayer feedforward network, trained error-back-propagation algorithm, is to estimate favourability gold deposits from raster GIS database Tenterfield 1:100000 sheet area, NSW (Australia). To validate assess effectiveness method, prospectivity maps...

In mining industry, accurate identification of new geographic locations that are favourable for mineral exploration is very important. However, definitive prediction such not an easy task. recent years, the use neural networks ensemble approach to classification problem has gained much attention. This paper discusses results obtained from using different network (NN) techniques prospectivtity problem. The proposed model uses information systems (GIS) data location. method tested on GIS...

Quantification of uncertainty in mineral prospectivity prediction is an important process to support decision making exploration. Degree uncertainly can identify level quality the prediction. This paper proposes approach predict degrees favourability for gold deposits together with quantification Geographic information systems (GIS) data applied integration ensemble neural networks and interval neutrosophic sets, three different network architectures are used this paper. The its represented...

This paper describes the integration of neural network ensembles and interval neutrosophic sets using bagging technique for predicting regional-scale potential mineral deposits as well quantifying uncertainty in predictions.