Search

Introduction The identification of potentially mineralized areas has progressed with the use and interpretation of all available exploratory data in the form of mineral potential modeling (MPM) (Yousefi and Nykänen, 2017). Recently, machine learning methods have been a popular research topic in the field of MPM ((Chen and Wu, 2016). Machine learning algorithms that have been used in MPM generally fall into the categories of being supervised or unsupervised. Supervised models, use the location of the known mineral occurrences as training sites (or labeled data). Therefore, these models suffer stochastic bias and error (Zuo and Carranza, 2011). Unsupervised models classify mineral prospectivity of every location based solely on feature statistics of individual evidential data layers ((Abedi et al., 2012). The semi-supervised learning models are a hybrid of supervised and unsupervised learning models that use both labeled and unlabeled data to extract the hidden structure of the data, as well as the relation between the input exploration layers and the output labeled data (Fatehi and Asadi, 2017). The Dehsalm study area forms a part of the Lut metallogenic block of eastern Iran, which is characterized by the subduction zone setting and extensive magmatism (Beydokhti et al., 2015). The objective of this research is to present a prospectivity model to delineate exploration target areas for porphyry Cu-Au mineralization in the study area. For generating a prospectivity model, we used TSVM algorithm, a semi-supervised learning integration technique, to identify the anomalous areas related to the porphyry Cu-Au mineralization. The input layers are selected based on a conceptual model for porphyry Cu-Au mineral system. The performance of the mineral prospectivity maps (MPMs) is evaluated using the various techniques, including the receiver operating characteristic (ROC) curve, an area under curve (AUC) metric. Materials and methods To apply a process-based understanding of porphyry copper-gold deposit system on the mapping of prospectivity, a conceptual model must be first developed (Fatehi and Asadi, 2017). Such a model should depict critical scale-dependent processes involved in the mineral deposit formation, and a mineral system approach can be followed to aid understanding where, when and why mineral deposits form (Parsa et al., 2016). The spatial data sets used to model porphyry Cu-Au prospectivity of the study area include geological, remote sensing, geophysical, and geochemical data. In this study, semi-supervised support vector machine (TSVM) prospectivity technique is utilized to model porphyry Cu-Au target areas. The TSVM is an extension of SVM that uses the unlabeled data to improve the performances of the classifier. The aim of TSVM algorithm is to find the decision hyper-plane subject to maximize the margin distance in labeled and unlabeled data. Result In the present study, TSVM and SVM models were applied to Cu-Au prospectivity modeling in the study area. The models were trained based on the location of known Cu-Au mineralization occurrences and non-deposit location using e1071 package in R open-source statistical software (70% of the labeled data were used in training and 30% in the testing phase of learning in both algorithms). The RBF kernel function were used and the optimal values of the kernel parameters were assessed using a 10-fold cross-validation procedure and the best learning performance was selected by correct classification rate. The output of the models highlighted the target areas for porphyry Cu-Au mineralization in the study area. The receiver operating characteristics (ROC) analysis shows that both models perform well, however, the TSVM model yields the best performance. Discussion To identify exploratory target areas on a regional scale, various supervised and unsupervised approaches have been developed in mineral potential modeling. Supervised methods such as SVM use labeled data to classify exploratory datasets. In this research, a new semi- supervised learning method, TSVM, was applied to model the mineral potential for porphyry Cu-Au mineralization in the Dehsalm exploration zone. The introduced target areas by TSVM method, within the known mineral indices, covered smaller areas than targets identified by SVM model, so planning the detailed exploration phase will be optimal. The result of this research demonstrates the superiority of the semi-supervised learning method in identifying the target areas for planning the exploratory operations. Acknowledgements We would like to thank the Geological Survey of Iran for providing the exploration data used in this research. The financial support of the South Khorasan Industry, Mine & Trade Organization is gratefully thanked. References Abedi, M., Norouzi, G.H. and Bahroudi, A., 2012. Support vector machine for multi-classification of mineral prospectivity areas. Computers & Geosciences, 46: 272–283. https://doi.org/10.1016/j.cageo.2011.12.014 Beydokhti, R.M., Karimpour, M.H., Mazaheri, S.A., Santos, J.F. and Klötzli, U., 2015. U-Pb zircon geochronology, Sr-Nd geochemistry, petrogenesis and tectonic setting of Mahoor granitoid rocks (Lut Block, Eastern Iran). Journal of Asian Earth Sciences, 111: 192–205. https://doi.org/10.1016/j.jseaes.2015.07.028 Chen, Y. and Wu, W., 2016. A prospecting cost-benefit strategy for mineral potential mapping based on ROC curve analysis. Ore Geology Reviews, 74: 26–38. https://doi.org/10.1016/j.oregeorev.2015.11.011 Fatehi, M. and Asadi, H.H., 2017. Data integration modeling applied to drill hole planning through semi-supervised learning: A case study from the Dalli Cu-Au porphyry deposit in the central Iran. Journal of African Earth Sciences, 128: 147–160. https://doi.org/10.1016/j.jafrearsci.2016.09.007 Parsa, M., Maghsoudi, A., Yousefi, M. and Sadeghi, M., 2016. Prospectivity modeling of porphyry-Cu deposits by identification and integration of efficient mono-elemental geochemical signatures. Journal of African Earth Sciences, 114: 228–241. https://doi.org/10.1016/j.jafrearsci.2015.12.007 Yousefi, M. and Nykänen, V., 2017. Introduction to the special issue: GIS-based mineral potential targeting. Journal of African Earth Sciences, 128: 1–4. https://doi.org/10.1016/j.jafrearsci.2017.02.023 Zuo, R. and Carranza, E.J.M., 2011. Support vector machine: a tool for mapping mineral prospectivity. Computers & Geosciences, 37(12): 1967–1975. https://doi.org/10.1016/j.cageo.2010.09.014

Introduction Fluid inclusions are small, usually microscopic, volumes of pore fluid, which are crystallographically trapped in rocks during diagenesis or fracture healing processes. Nowadays, various techniques are used for resource exploration. Application of a fluid inclusion is one of these methods that has been developed for mineral, geothermal, and petroleum reservoir exploration. The study of fluid inclusions represents our most reliable source of information on the temperature, pressure, and fluid composition data of the ore fluid, and it is one of the most important tools for research into the economic geology and genesis of ore deposits (Moon, 1991). To achieve these goals, transparent and polished slabs of rock material are prepared and optically studied with a petrographic microscope. Samples are viewed under transmitted plane-polarized white light as well as under reflected ultraviolet or blue-violet illumination. During the fluid inclusion petrography, the volume fractions of phases are routinely estimated at room temperature to deduce whether assemblages of cogenetic inclusions were originally trapped from a one-phase or a multi-phase pore fluid. In the present research study, the microthermometric properties of the fluid inclusion data through pressure, temperature, and salinity diagrams were computed by geometrical modeling of fluid inclusion (Bakker and Larryn, 2006). The proposed method provides a quick and low cost technique to preliminarily investigate the microthermometric parameters of the fluid inclusion. To evaluate the proposed geometrical model, the Mehdiabad Pb-Zn deposit is selected as the case study. The Mehdiabad Pb-Zn deposit is located at the Yazd-Anarak metallogenic belt, 110 km southeast of Yazd, in the Central Iran structural zone. The host rocks of the deposit consist of lower Cretaceous silty limestone and dolomite. The main occurrences are the Calamine mine (CM), the Black-Hill ore (BHO), the East Ridge (ER) and the Central Valley Orebody (CVOB). The ore body consists of a primary sulfide ore and a supergene non-sulfide ore, the latter one having been mined at CM (Ghasemi, 2007; Rajabi et al., 2012). Materials and methods The shape and geometry of fluid inclusion are one of the most important parameters, which were applied to estimate 3D degree of filling and find the useful information about temperature, pressure, salinity and depth of trapping without using time-consuming and costly heating-cooling operation. Inclusions in normal thick-sections are rotated stepwise and their projected areas and area-fractions are plotted against rotation angle. The outputs are systematically related to inclusion orientation, inclusion shape, and filling degree. The dependency on orientation is minimized when area fractions are measured at the position where the inclusions project their largest total areas. The shape factor is employed to present a new objective classification of inclusion projections, based on the extracted parameters from digital image processing (Bakker and Larryn, 2006). In this research, Mehdiabad Pb-Zn deposit has been chosen to evaluate the proposed method. Based on the fluid inclusion petrography, four fluid inclusion types are observed: 1) L+V; 2) L+L; 3) L; and 4) V; L+V phase is the most popular. After preparing 2D image of sections, 2D and 3D degree of fills were calculated by measuring the areas of total, bubble, and spot of fluid inclusion and computing the third dimension (Z) of fluid inclusion. Four geometrical models of volume fractions are defined, including cylinder, tetragonal prism, truncated cone, hexagon, and ellipsoid (Bakker and Larryn, 2006; Hossein Morshedy et al., 2008). In this case study, 3D proper models of the fluid inclusions are selected, depending on its geometry (hexagonal or ellipsoid). Then 2D degrees of filling (area fraction) is converted to 3D degrees of filling (volume fraction). The geometrical modeling results are well matched with computational outputs. Results and discussion In this research, the ratios of area to volume fractions in geometrical and computational modelling were calculated 0.75 and 0.77, respectively. In the Mehdiabad Pb-Zn deposit, the main geometrical shapes of fluid inclusions were followed up the hexagonal prism with hexagonal pyramids and ellipsoid models. 3D geometrical modeling of fluid inclusion showed vapor fraction, 25% and density, 0.7 g/cm3, which the microthermometric and other parameters were obtained homogenization temperature nearly 100-200 °C (average of 150 °C), pressure between 400-500 ATM, formation temperature about 250-350 °C, salinity within a range of 10 to 15 wt.% NaCl equiv. and depth of mineralization 150-200 m. This finally achieved results have a high similarity with the typical carbonate-hosted Pb-Zn deposit. References Bakker, R.J. and Larryn, W.D., 2006. Estimation of volume fractions of liquid and vapor phases in fluid inclusions, and definition of inclusion shapes. American Mineralogist, 91(1): 635–657. Ghasemi, M., 2007. Genesis of Mehdiabad Pb–Zn deposit and comparing with other Pb–Zn deposits. M.Sc. Thesis, Research Institute for Earth Science, Geological Survey of Iran, Tehran, Iran, 238 pp. (in Persian with English abstract) Hossein Morshedy, A., Mojtahedzadeh, H. and Kohsary, A.H., 2008. Measuring microthermic parameters of fluid inclusion with studying their geometries and models, case study: Mehdiabad Pb–Zn deposit. 15th Symposium of Crystallography and Mineralogy of Iran, Ferdowsi University of Mashhad, Mashhad, Iran. (in Persian with English abstract) Moon, K.J., 1991. Application of fluid inclusions in mineral exploration. Journal of Geochemical Exploration, 42(1): 205–221. Rajabi, A., Rastad, E. and Canet, C., 2012. Metallogeny of Cretaceous carbonate-hosted Zn–Pb deposits of Iran: geotectonic setting and data integration for future mineral exploration. International Geology Review, 54(14): 1649–1672.

Introduction Nowadays, exploration of rare earth element (REE) resources is considered as one of the strategic priorities, which has a special position in the advanced and intelligent industries (Castor and Hedrick, 2006). Significant resources of REEs are found in a wide range of geological settings, including primary deposits associated with igneous and hydrothermal processes (e.g. carbonatite, (per) alkaline-igneous rocks, iron-oxide breccia complexes, scarns, fluorapatite veins and pegmatites), and secondary deposits concentrated by sedimentary processes and weathering (e.g. heavy-mineral sand deposits, fluviatile sandstones, unconformity-related uranium deposits, and lignites) (Jaireth et al., 2014). Recent studies on various parts of Iran led to the identification of promising potential of these elements, including Central Iran, alkaline rocks in the Eslami Peninsula, iron and apatite in the Hormuz Island, Kahnouj titanium deposit, granitoid bodies in Yazd, Azerbaijan, and Mashhad and associated dikes, and finally placers related to the Shemshak formation in Marvast, Kharanagh, and Ardekan indicate high concentration of REE in magmatogenic iron–apatite deposits in Central Iran and placers in Marvast area in Yazd (Ghorbani, 2013). Materials and methods In the present study, the geochemical behavior of rare earth elements is modeled by using multivariate statistical methods in the eastern part of the Marvast placer. Marvast is located 185 km south of the city of Yazd in central Iran between Yazd and Mehriz. This area lies within the southeastern part of the Sanandaj-Sirjan Zone (Alipour-Asll et al., 2012). The samples of 53 wells were analyzed for Whole-rock trace-element concentrations (including REE) by inductively coupled plasma-mass spectrometry (ICP-MS) (GSI, 2004). The clustering techniques such as multivariate statistical analysis technique can be employed to find appropriate groups in data sets. One of the main objectives of data clustering is to maximize both the similarity within each cluster and the difference between clusters, and finally find the structure in the data. Nowadays, cluster analysis is applied in many disciplines: biology, botany, medicine, psychology, geography, marketing, image processing, psychiatry, archaeology, etc. (Everitt et al., 2011). To execute a partitioning algorithm, the principal components analysis (PCA) algorithm is applied for feature selection, feature extraction and dimension reduction. Hierarchical clustering can be utilized to provide a nested sequence of partitions with bottom-up or top-down methods based on similarity. The single linkage and complete linkage are the most popular hierarchical algorithms (Jain et al., 1999; Ji et al., 2007). Results and discussion The REE chondrite-normalized pattern for the eastern area in the Marvast placer represents a high match to the standard pattern of monazite. This pattern shows the positive anomaly of Ce and the negative anomaly of Eu. To determine the distribution of REEs concentration, 2D interpolation maps were plotted in three groups of light, middle, and heavy REEs (LREE, MREE, and HREE), which were indicated in the geochemical anomaly at the south and south-west of the area. The relative ratios of (LREE/HREE) and (Ce/Eu) exposed the high proportion of LREEs to HREEs. In the next section, the hierarchical clustering algorithm was employed to partition the data in the feature and sample levels. The elements portioning demonstrated four separated groups, which can be related to atomic and chemical structures. The studied region was divided into four zones by the clustering approach. The fourth zone confine coincided with the REE anomaly area. Finally, PCA was applied as the multivariate statistical tool to this dataset. Hence three principal components modeled over 90% of the variance. For the first component, the distribution map of load factor has a good agreement with anomaly area. References Alipour-Asll, M., Mirnejad, H. and Milodowski, A.E., 2012. Occurrence and paragenesis of diagenetic monazite in the upper Triassic black shales of the Marvast region, South Yazd, Iran. Mineralogy and Petrology, 104(3-4): 197-210. Castor, S.B. and Hedrick, J.B., 2006. Rare earth elements. In: J.E. Kogel (Editors), Industrial minerals & rocks: commodities, markets, and uses. Society for mining, metallurgy, and exploration, Inc. (SME), Colorado, pp. 769-792. Everitt, B., Landau, S., Leese, M. and Stahl, D., 2011. Cluster Analysis. 5th ed., Wiley Publishing, Hoboken, Geological Survey of Iran (GSI), 2004. Exploration of rare earth element and monazite in the alluvium of the southern Marvast (eastern area). Tehran, 66 pp. (in Persian) Ghorbani, M., 2013. The economic geology of Iran: mineral deposits and natural resources. Springer, London, 569 pp. Jain, A.K., Murty, M.N. and Flynn, P.J., 1999. Data clustering: a review. Association for Computing Machinery computing surveys, 31(3): 264-323. Jaireth, S., Hoatson D.M. and Miezitis Y., 2014. Geological setting and resources of the major rare-earth-element deposits in Australia. Ore Geology Reviews, 62: 72-128. Ji, H., Zeng, D., Shi, Y., Wu, Y. and Wu, X., 2007. Semi-hierarchical correspondence cluster analysis and regional geochemical pattern recognition. Journal of Geochemical Exploration, 93(2): 109-119.

Introduction Mohammad-abad Oryan is the only potential source of borate in the North-east of Iran located in 50 km South of Sabzevar. The area is located in tuff marl, tuffaceous marl, volcanic braccia and tuff braccia structures. Remote sensing techniques, geological studies and integration of this data in GIS were applied in an area of about 600 square kilometers to locate the promising areas of borate mineralization for detailed studies (Bemani, 2012). The aim of this detailed geochemical study is to confine the anomaly areas for exploratory drilling and trenching. Materials and methods Field studies were carried out in 9 geological traverses, mainly in Tonakar and Borje Kharkan area and 126 brine samples were taken from hydrothermal springs and 13 rock samples were taken from trenches. All the samples were analyzed for four elements, including B, K, Li and Mg. In order to determine the threshold quantities of the samples and isolation of anomaly, the data were analyzed using statistical methods including classical statistics, fractal geometry and EDA methods (Bemani, 2012). Result Initial data analysis showed that there were no censored data. Also, by applying statistical hypothesis testing, no significant relation was observed between the elements in the two areas (except for Li). Therefore, all the statistical analyses were carried out separately. After outlier correction, based on the amount of skewedness and histograms and probability plots of different elements, it became clear that none of the elements in the raw data distribution were normal and required to be transformed to be close to normal. In this study, logarithmic and three-parameter logarithm transformation were used in order to normalize the data . Based on the mean values, standard deviation of the normalized data, and background value and threshold, probable and possible anomalies were obtained and geochemical anomaly maps were drawn to identify the promising areas. With the exception of the fractal pattern, anomaly separation methods are based on the differences of fractal dimensions between communities of geochemical data (Hasanipak and Sharafoddin, 2005). In this study, concentration area fractal method was used to separate anomalies from the background. Using fractal geometry, threshold value corresponding to the two areas (Tonakar and Borje Kharkan) were obtained and were plotted separately on geochemical maps. Exploratory data analysis (EDA) is an approach to analyze data sets to summarize their main characteristics, often with visual methods (Filliben and Heckert, 2005). Exploratory data analysis is a useful method for analysis of geochemical exploration data. This is a statistical method known as the Robust Statistic classification (Carranza, 2009). In geochemical exploration, box plots, histograms and scatter plot are more practical. According to the box plots, the data of Tonakar and Borje Kharkan areas were classified and threshold levels were determined (Bemani, 2012). Discussion Using the results obtained from different methods, geochemical maps of each area were prepared for all the elements and thresholds were obtained for each method. Moreover, the geochemical maps of each area were plotted for each element. According to the geochemical maps of Tonakar area, boron anomaly was observed in the East and West zones and the anomaly of the latter is larger. These areas were recommended for further detailed exploration and borehole drilling. Also, geochemical maps of Borje Kharkan showed anomaly in the central zone for all of the elements. The results showed that the highest and the lowest amounts of boron in brines samples vary between 6 ppm to 5930 ppm. Among boron and the three other elements (i.e. lithium, magnesium and potassium) a significant correlation was not observed. In terms of frequency, in most cases brines with high levels of boron (more than 1000 ppm) were concentrated in the South East of the Tonakar area. So, this area was suggested for detailed exploration (Bemani et al., 2014). Generally speaking, for considering the spatial distribution of data the fractal method could better identify the anomalies. Also, EDA is a quick and easy method to detect anomalies. Acknowledgment The authors are grateful to the Kaniran Mining Company and the south Khorasan branch of the Iranian Mining Engineering Organization for their financial support of this study. References Bemani, M., 2012. Prospecting and Exploring of Borax in the south of Sabzevar, combination of remote sensing, field surveying and geochemical studying. M.Sc. Thesis, University of Yazd, Yazd, Iran, 137 pp (in Persian with English abstract). Bemani, M., Mojtahedzadeh, S.H. and Kohsari, A.H., 2014. Investigation of geology, mineralogy and genesis of Mohammadabad-Oryan index boron (south of Sabzevar). Iranian Journal of Crystallography and Mineralogy 22(1): 173- 186 (in Persian with English abstract). Carranza, E.J.M., 2009. Geochemical Anomaly and Mineral Prospectivity Mapping in GIS. In: M. Hale (Editor), Handbook of exploration and environmental geochemistry. Elsevier, Amsterdam, pp. 51-115. Filliben, J.J. and Heckert, A., 2005. Exploratory data analysis. Engineering Statistics Handbook, Internet, National Institute of Standards and Technology, http://www.itl.nist.gov/div898/handbook/eda/section3/eda356.htm. Hasanipak, A.A. and Sharafaddin, M., 2005. Exploratory Data Analysis.Tehran University press, Tehran 996 pp (in Persian).