Currently, traditional resource estimation techniques i.e., Kriging, inverse distance square or conditional simulation are commonly used as best practice to characterise the mineral deposit. However, due to the formation of the ore involves complex geological process that are often nonlinear and nonstationary, accurate prediction of mineral grade distribution can be difficult and inaccurate using these traditional techniques. Machine Learning (ML) methods can offer valuable support and\/or alternatives for resource estimation by complementing traditional techniques. Methods such as neural networks, gradient boosting or hybrid-based methods can be used to validate the estimation from traditional techniques and increase the accuracy of the estimation. The accuracy of ML methods can be further improved by incorporating not only the geological coordinates system of training data points but also additional geological features such as rock type, lithology variation and alteration information. Moreover, integrating geophysical information into the training dataset can further increase the precision of predictions. In addition, as the new data becomes available with mining, ML models can be continuously updated, enabling dynamic and increasingly reliable estimations for the mining industry. This research proposes the development of ML methodologies and framework for the mineral resource estimation as a robust alternative method to traditional resource estimation methods.<\/p>\n\n\n\n
Aims<\/p>\n\n\n\n
The aim of this research is to offer a paradigm shift in resource estimation practice in mining projects. It aims to propose a new methodology based on Machine learning models to estimate mineral resources that enhances the accuracy and reliability of grade predictions by integrating not only the distance-based coordinate system but also geological and geophysical data, offering a dynamic and adaptable alternative to traditional estimation techniques.<\/p>\n\n\n\n
Objectives<\/p>\n\n\n\n
\u2022 To integrate additional geological and geophysical information\u2019s into ML prediction models to increase the accuracy of the grade estimation.
\u2022 To compare the estimation results from ML models vs traditional techniques using industrial dataset.<\/p>\n\n\n\n
Significance<\/p>\n\n\n\n
Mineral resource estimation is an important stage of mining project which impacts the entire value chain of the process. Traditional estimation methods often fail to estimate accurate mineral grades especially complex and variable geological domain which directly impacts the viability of the mining project. This research introduces a new framework using ML techniques to improve the accuracy of the grade estimation further geological and geophysical features. ML can capture not only nonlinear relationships and patterns that conventional methods may overlook but also can learn from new data which allows for continuous model refinement, leading to more dynamic and accurate estimations over time.<\/p>\n\n\n\n
Ideal Candidate<\/p>\n\n\n\n
In order to carry out this research, candidate should possess a strong knowledge on geoscience, data science, and ML modelling. Candidates can hold a degree in Geology \/ Geological Engineering, Mining Engineering, Geophysics, Data Science, Computer science, Applied mathematics\/statistics. Additionally, the applicants should meet the eligibility criteria for entry into a PhD program at Curtin University. This project is open to International and Domestic applicants.<\/p>\n\n\n\n
Scholarship<\/p>\n\n\n\n
If you are identified as the preferred candidate for this project, you may be considered for an RTP scholarship<\/a>.<\/p>\n\n\n\n Enquires and How to Apply<\/p>\n\n\n\n For enquires about this opportunity contact Professor Erkan Topal at E.Topal@curtin.edu.au<\/a><\/p>\n\n\n\n