Orogenic gold prospectivity mapping using machine learning
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
SummaryAs major mineral discoveries have become rarer over the last two decades, the industry has begun to turn to new technologies to assist in the exploration process. One such advancement is the application of machine learning and artificial intelligence (AI) to geoscience data. Mineral prospectivity mapping has been around for decades but with the increase in computer power, recently it has gained traction again as a means for exploration teams to take full advantage of the numerous datasets at their disposal. Although having a team of human experts with a wealth of geoscience knowledge and experience is still fundamental to the exploration process, the ability to robustly integrate and analyse large geoscience datasets over vast spatial regions quickly becomes unwieldy if done manually.In this study, we developed a new algorithm for mineral prospectivity mapping using a VNet deep convolutional neural network and applied it to finding gold at the Committee Bay greenstone belt in the Canadian Arctic. The machine learning network took all the geoscience data available from the area and generated a prospectivity map for targeting economic orogenic gold mineralization. The results were subsequently validated on a separate nearby region where the machine predictions were compared to gold assay values from drilling. The gold assays from this region were not included in the training process, and the method demonstrated good success in predicting where the highest gold mineralization occurred.A subsequent gold prospectivity map was produced for the main area in question, and in addition to many new targets the VNet algorithm predicted many targets that the exploration team had previously generated. This suggests that this process assists the exploration team in vetting old targets while opening their eyes to new targets as well. In this way, the algorithm helps to vector in on prospective new and old areas while maximizing the value of all available geoscience data.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it