A Machine Learning Zircon Trace Element Tool to Predict Porphyry Deposit Type and Resource Size
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
Abstract Porphyry deposits are primarily known for their association with base metals like copper and to some extent molybdenum and gold. Here we present machine learning models, based on zircon composition, that provide quantitative distinction between different deposit types and resource sizes. Using a global zircon compositional database for different porphyry deposits (9,649 samples), we trained several machine learning models. A porphyry deposit type model (PDT model) was developed using XGBoost, which distinguishes between barren, Cu, and Mo bearing deposits. Furthermore, porphyry Cu and Mo reserve models (Porphyry Cu Reserve [PCR] and Porphyry Mo Reserve [PMR] model) were also developed using XGBoost and LightGBM, respectively, to give prediction of resource size in unexplored area. F1‐scores for the models are 0.97, 0.91, and 0.82. The model‐built feature importance and Shapley Additive exPlanations values imply that (Eu N /Eu N *)/Y, Th/U, Th/U and Ce are important in the PDT model, Ti, T (°C), U, and Hf are important for the PCR model, and Hf, U, Th/U, and Eu N /Eu N * are important for the PMR model. From a mineral system perspective, the three models imply that water, temperature, and magma evolution are pivotal to the type of deposits that forms. Temperature and magma evolution in particular are important in prediction of Cu and Mo resource size. Application of models to the Wunugetushan deposit gives ore type and resource predictions that are consistent with known deposit occurrence and geochemistry. These findings suggest that machine learning models may not only assist in understanding the main geological processes linked to porphyry mineralization, but also have application in reducing exploration risk.
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 | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.002 |
| 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