Using machine learning to estimate a key missing geochemical variable in mining exploration: Application of the Random Forest algorithm to multi-sensor core logging data

Schnitzler, Nathalie; Ross, Pierre-Simon; Gloaguen, Erwan

doi:10.1016/j.gexplo.2019.106344

Cited by 28 publications

(10 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Geological mapping [22][23][24] studies were conducted using the characteristics of rocks. A mineral analysis [25][26][27][28][29][30][31] was conducted using drilling data or samples. A mineral prospectivity modeling and mapping [32][33][34][35][36][37][38][39][40][41][42] study was performed to evaluate the potential of minerals using the exploration data.…”

Section: Publication Sourcementioning

confidence: 99%

Systematic Review of Machine Learning Applications in Mining: Exploration, Exploitation, and Reclamation

Jung¹,

Choi²

2021

Minerals

View full text Add to dashboard Cite

Recent developments in smart mining technology have enabled the production, collection, and sharing of a large amount of data in real time. Therefore, research employing machine learning (ML) that utilizes these data is being actively conducted in the mining industry. In this study, we reviewed 109 research papers, published over the past decade, that discuss ML techniques for mineral exploration, exploitation, and mine reclamation. Research trends, ML models, and evaluation methods primarily discussed in the 109 papers were systematically analyzed. The results demonstrated that ML studies have been actively conducted in the mining industry since 2018, mostly for mineral exploration. Among the ML models, support vector machine was utilized the most, followed by deep learning models. The ML models were evaluated mostly in terms of their root mean square error and coefficient of determination.

show abstract

Section: Publication Sourcementioning

confidence: 99%

Systematic Review of Machine Learning Applications in Mining: Exploration, Exploitation, and Reclamation

Jung¹,

Choi²

2021

Minerals

View full text Add to dashboard Cite

show abstract

“…The results show that RF is a better technique for compositional discrimination in the Broken Hill domain compared with linear discriminant analysis. In an exploration study, Schnitzler et al [113] used RF to assess sodium (Na) concentration in the Matagami mining district of Québec, Canada and illustrated that RF could be an efficient tool for estimating missing or unmeasured geochemical elements in an exploration database. Matin and Chelgan [114] also alluded to the potential of RF to model complex relationships in their study to investigate gross calorific value of coal samples from 26 US states, where it was applied with satisfactory results.…”

Section: Random Forest (Rf)mentioning

confidence: 99%

Machine Learning—A Review of Applications in Mineral Resource Estimation

Dumakor-Dupey

Arya

2021

Energies

View full text Add to dashboard Cite

Mineral resource estimation involves the determination of the grade and tonnage of a mineral deposit based on its geological characteristics using various estimation methods. Conventional estimation methods, such as geometric and geostatistical techniques, remain the most widely used methods for resource estimation. However, recent advances in computer algorithms have allowed researchers to explore the potential of machine learning techniques in mineral resource estimation. This study presents a comprehensive review of papers that have employed machine learning to estimate mineral resources. The review covers popular machine learning techniques and their implementation and limitations. Papers that performed a comparative analysis of both conventional and machine learning techniques were also considered. The literature shows that the machine learning models can accommodate several geological parameters and effectively approximate complex nonlinear relationships among them, exhibiting superior performance over the conventional techniques.

show abstract

“…Similarly, fine-resolution multi-parameter datasets obtained from drill cores cannot be used effectively in isolation without considering the large-scale structural architecture of the deposit. Current and future machine learning (ML) systems that maximize the knowledge gained from drill core data (Bérubé et al 2018;Liu et al 2019;Schnitzler et al 2019) hold great promise for the minerals industry. Even though the benefits of ML are great, it is paramount that geologists know how to train these artificial computer algorithms to recognize large-scale hierarchical structural controls, otherwise ML systems may never yield sensible exploration targets.…”

Section: Implications For Orogenic Gold Mineralization and Explorationmentioning

confidence: 99%

Deposit-scale structural architecture of the Sigma-Lamaque gold deposit, Canada—insights from a newly proposed 3D method for assessing structural controls from drill hole data

Cowan

2020

Miner Deposita

View full text Add to dashboard Cite

Deposit-scale structural analysis is a practical method of structurally analyzing 3D drilling data. This interpretation method uses readily available distributed 3D data at the most relevant scale for mining and exploration. This method does not depend on collecting traditional structural data from core, such as orientation of planar and linear data, but relies on the premise that grade distributions sampled from drilling provide insight into the structural control of the deposit, with the interpretation of structural geometry done using a computer 3D viewer. This method is applied to the well-studied Archean Sigma-Lamaque gold deposit, which features late-orogenic vein mineralization, and which is controlled by three orders of structural features. The first-order control is the easterly plunging pipe-like conduit parallel to F 2 fold plunge defined by the intersection of bedding (S 0 ) and foliation (S 2 ) (~1.5 km radius × 3 km down-plunge). The second-order controls, nested within the first-order, are the younger subvertical Late Diorite plugs that intruded vertically into the host rocks along the established S 2 foliation anisotropy (1250:400:150 m dimension for Main Plug). The third-order control, laterally limited by first-and second-order structural controls, is the well-documented goldbearing vein sets (1-10 m thick and up to 1.5 km in extent) that developed in the Late Diorites and in the older folded volcanogenic stratigraphy and diorites. It is proposed that the first-order control defined by the moderately plunging F 2 fold is the main fluid pathway resulting from structural permeability formed during earlier folding. The second-order Late Diorite plugs and other competent lithological units, which intersected and fractured during late-stage D 2 compression, served as chemically reactive sites causing gold to precipitate along this first-order pathway. It is predicted that intrusions that do not intersect the first-order structural conduits are less likely to be mineralized. Greenfield orogenicgold exploration requires that the first-order controls are identified, and the geometry of hydrothermal alteration is best understood in the context of the structural architecture of the host rocks.

show abstract

Using machine learning to estimate a key missing geochemical variable in mining exploration: Application of the Random Forest algorithm to multi-sensor core logging data

Abstract: E (2019) Using machine learning to estimate a key missing geochemical variable in mining exploration: application of the Random Forest algorithm to multi-sensor core logging data.

Cited by 28 publications

References 21 publications

Systematic Review of Machine Learning Applications in Mining: Exploration, Exploitation, and Reclamation

Systematic Review of Machine Learning Applications in Mining: Exploration, Exploitation, and Reclamation

Machine Learning—A Review of Applications in Mineral Resource Estimation

Deposit-scale structural architecture of the Sigma-Lamaque gold deposit, Canada—insights from a newly proposed 3D method for assessing structural controls from drill hole data

Contact Info

Product

Resources

About