A Comparative Study of Machine Learning and Traditional Techniques for Grade Prediction and Grade-Tonnage Evaluation in a Small VMS Deposit

Aug 4, 2026

Upload by Aurelia

Small-scale, high-grade volcanogenic massive sulfide (VMS) deposits challenge resource estimation because of strong grade variability and complex geology. This thesis tests whether machine learning improves gold grade prediction and grade-tonnage evaluation versus traditional geostatistical and distance-based methods. A 3D block model (5×5×5 m) in Vulcan is evaluated using IDW, Simple Kriging, Ordinary Kriging, and ensemble tree models, with cross-validation versus independent train-test testing. Results show higher test accuracy for machine learning, with low-information noise mitigated by filtering Au > 0.0001 g/t. Spatial comparisons, residuals, and grade-tonnage curves reveal method-specific behaviors and improved discrimination and preservation of localized high-grade zones after filtering.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages68

Reading Time2 h 51 min

AccessFree to download

Abstract

General Audience Abstract

Method Comparison: IDW, Kriging, and Ensemble Tree Models

Filtering and Geologically Realistic Tonnage

Spatial Comparisons and Residual Analysis

Conclusions and Practical Implications

Acknowledgements

A Comparative Study of Machine Learning and Traditional Techniques for Grade Prediction and Grade-Tonnage Evaluation in a Small VMS Deposit Cemile Dilara Bag Thesis submitted to the faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of Master of Science In Mining and Minerals Engineering Erik Westman, Chair Ben M. Frieman Rohit Pandey December 5, 2025 Blacksburg, VA Keywords: machine learning; ore grade estimation; block modelling; grade-tonnage curve; VMS deposit; Random Forest; Gradient Boosting, IDW; Kriging A Comparative Study of Machine Learning and Traditional Techniques for Grade Prediction and Grade-Tonnage Evaluation in a Small VMS Deposit Cemile Dilara Bag ABSTRACT Small-scale, high-grade volcanogenic massive sulfide (VMS) deposits present unique challenges for resource estimation due to their strong grade variability and complex geological structures. This thesis evaluates whether machine learning methods can improve grade prediction and tonnage estimation compared to traditional methods. A threedimensional block model with 5 x 5 x 5 m resolution was constructed in Vulcan, and grade estimation was performed using Inverse Distance Weighting (IDW), Simple Kriging (SK) , Ordinary Kriging (OK), and ensemble tree models. Traditional methods were assessed using cross-validation within Vulcan, while machine-learning models were evaluated using an independent train-test split. Approximately six million block centroids were exported for full model prediction to compare all methods directly. Machine learning models produced the highest accuracy in the test set but generated low-level noise predictions across sparsely informed areas. A filtering threshold of Au > 0.0001 g/t was applied to mitigate this effect and achieve geologically realistic tonnage estimates. Spatial block-model comparisons, residual analyses, and grade-tonnage curves showed distinct behaviors among methods. IDW yielded the highest tonnage at low cutoffs, Simple Kriging and Random Forest exhibited similar behavior in sparsely informed areas, and Ordinary Kriging consistently produced conservative tonnage estimates. After filtering, ensemble machine learning models provided improved grade discrimination and preserved localized high-grade zones more effectively than traditional methods. This study demonstrates that machine learning approaches can complement traditional methods and offer enhanced performance for small VMS deposits. The results highlight practical considerations for applying machine learning in early-stage resource evaluation and emphasize the need for domain-based modeling in later stages. A Comparative Study of Machine Learning and Traditional Techniques for Grade Prediction and Grade-Tonnage Evaluation in a Small VMS Deposit Cemile Dilara Bag GENERAL AUDIENCE ABSTRACT Mineral deposits that contain gold and other valuable metals are becoming harder to find, especially the large deposits that historically supplied much of the world’s production. Asa result, smaller but high-grade deposits, such as those formed by ancient underwater volcanic activity, are becoming increasingly important. Estimating how much metal these deposits contain is a key step in planning a mine, but this is challenging when drillhole data are limited and the rocks vary greatly over short distances. This thesis explores whether modern computer-based approaches, known as machinelearning methods, can improve these estimates compared with traditional techniques used by geologists. Using a three-dimensional model of a gold-rich volcanic deposit, several estimation methods were tested and compared. Traditional methods rely mainly on distance and spatial patterns, while machine-learning methods learn relationships directly from the data. The results show that machine-learning methods can provide more accurate predictions where drilling information is available,

FAQs

What problem does the thesis address in small VMS deposits?

It addresses resource estimation challenges caused by strong grade variability and complex geological structures, especially when drillhole data are limited.

Which estimation methods are compared for grade and tonnage evaluation?

Traditional methods include IDW, Simple Kriging, and Ordinary Kriging, while machine learning uses ensemble tree models evaluated with a train-test split.

How does the thesis handle unrealistic noise predictions from machine learning?

It applies a filtering threshold of Au > 0.0001 g/t to reduce low-level noise and produce geologically realistic tonnage estimates.

Aug 4, 2026

Upload by Aurelia