Rocks, Data, Algorithms: A Roadmap for Practical Machine Learning Adoption in Rock Engineering
Bibliographic record
Abstract
ABSTRACT: Recent advancements in sensor technologies, computing and data storage, and accessibility to commercial AI tools has resulted in an increase in machine learning in rock engineering research. Many researchers have been able to demonstrate data-driven algorithms that have the potential to transform rock engineering paradigms. These studies include developing ML for predicting geology with high accuracy, forecasting rock bursts in high stress environments, and decreasing the computational time needed to update complex numerical models. Despite the potential shown in the research literature, machine learning (ML) has not yet crossed the threshold into practical use in rock engineering. This delay in adoption can be seen as an opportunity to standardize ML development in rock engineering to avoid some of the common pitfalls that can result in erroneous or nonsensical predictions, prior to widespread adoption. Namely, typical rock engineering projects suffer from three primary problems that will affect the trust rock engineers have in using ML: 1) Biased/poor quality data resulting in biased ML models 2) Insufficient data quantity resulting in unreliable ML models 3) Misunderstanding the trained ML model resulting in opaque rock engineering decisions This presentation will highlight the source of these methodological issues by illustrating them with examples and will discuss possible strategies to avoid them when applying ML to rock engineering.
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 | 0.059 | 0.096 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.004 |
| Science and technology studies | 0.001 | 0.007 |
| Scholarly communication | 0.009 | 0.018 |
| Open science | 0.005 | 0.006 |
| Research integrity | 0.008 | 0.010 |
| Insufficient payload (model declined to judge) | 0.013 | 0.005 |
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".