Comparative machine learning analysis for gold mineral prediction using random forest and XGBoost: A data-driven study of the Greater Bendigo Region, Victoria
Bibliographic record
Abstract
Gold mineral exploration remains critical to supporting global industries, yet traditional methods relying on manual interpretation of geophysical data are increasingly inefficient and prone to error, particularly when targeting undercover deposits. In Australia, most exploration research has focused on Western Australia, while the Greater Bendigo region in Victoria remains underexplored using modern data-driven approaches, despite its rich mining history and availability of high-resolution geophysical datasets. This study aims to demonstrate that a geospatial analysis methodology based on a machine learning approach enables high-accuracy prediction of gold mineral deposits in Bendigo. The methodology integrates geophysical data, including gravity, total magnetic intensity, and radiometric surveys, combined with geospatial preprocessing, scalable multi-resolution modelling, spatial labelling, and ensemble machine learning techniques, using Random Forest as the primary algorithm and XGBoost as a comparative model. Model performance was assessed using accuracy scores, ROC-AUC metrics, and spatial validation methods, including checkerboard and cluster-based cross-validation, across different spatial scales. Results showed that gravity and magnetic features were the strongest predictors, while radiometric features provided supporting information. Coarser spatial resolutions produced more stable predictions, reflecting regional geological patterns. The study presents a reproducible and adaptable machine learning methodology that addresses key exploration challenges and advances mineral prospectivity analysis using open-access geophysical data. • Machine learning applied to predict gold mineralisation in Greater Bendigo, Victoria. • Integrated gravity, magnetic, and radiometric data with geospatial preprocessing. • Random Forest used as the primary model with XGBoost for comparative analysis. • Model performance evaluated using ROC-AUC through checkerboard and cluster-based validation. • Established a reproducible workflow for spatially informed mineral prospectivity mapping.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".