Machine Learning-based Mineral Prospectivity Mapping: Exploring the Role of Negative Training Labels to Enhance Predictive Models
Bibliographic record
Abstract
Mineral Prospectivity Mapping (MPM) is an important tool to identify areas with significant potential to host mineral deposits. Recent advancements in computational sciences, especially the advent of Machine Learning (ML), have enhanced MPM's capabilities. ML techniques enable a higher degree of data integration and extraction compared to traditional statistical methods such as Weights of Evidence, enhancing the accuracy and efficiency of identifying mineral exploration zones. When using ML techniques for MPM, the influence of negative training labels (ie. barren areas with no mineralization) remains a neglected research area, and this study investigates the influence of such label selection to optimize predictive models for Canadian critical mineral exploration. Previous approaches to ML-based MPM often adopted a random assignment of negative training labels wherever positive training labels were absent. This study aims to refine this method, striving for a more systematic approach in negative label selection. The evolution of MPM, transitioning from traditional statistical methods to modern ML algorithms, signifies a shift towards heightened accuracy and efficiency. Prior research underscores the importance of balanced representation between mineralized and non-mineralized labels in ML models. Techniques such as Synthetic Minority Over-Sampling (SMOTE) and Positive and Unlabelled Learning (PUL) have been highlighted in previous studies, emphasizing the necessity of effectively handling negative training labels to prevent biases and enhance model performance. While SMOTE and PUL synthetically balance datasets by either oversampling minority classes or considering only positive and unlabeled instances, this study focuses on leveraging public exploration data to identify real negative training labels and provide a more authentic representation of non-mineralized areas without synthetic augmentation. Using datasets compiled by the Geological Survey of Canada containing discoveries & occurrences of magmatic Ni (±Cu ±Co ±PGE), this research incorporates geological, geochemical, and geophysical data from established sources. Public exploration data will be used to identify areas devoid of magmatic Ni (±Cu ±Co ±PGE). These locations will serve as negative training labels for this study. Our particular choice of ML model is a Gradient Boosting Machine (GBM), and validation involves comprehensive evaluation techniques such as confusion matrices and receiver operating characteristic curves to assess model performance.
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 | 0.006 | 0.018 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".