A Comparison of Lithology Predictors in Some Thin Bedded Gas Turbidite Reservoirs Using Conventional and Quantum Neural Networks
Bibliographic record
Abstract
Distinguishing between different lithologies is an important component of reservoircharacterization. It is particularly important in thin bedded gas turbidite reservoirs, where most of the gas isoften located in thin sand layers. If one has core material then identifying the various lithologies can usuallybe relatively straightforward. However, core retrieval is expensive, so cores are generally only obtainedfrom a small fraction of drilled wells. Thus, lithology profiles need to be estimated from other data such aswell logs. The present study compared different neural network approaches to predict lithology in thinbedded gas turbidite reservoirs in two wells in different regions: one in the Nile delta and the other inMiocene sediments in the Polish Carpathian Foredeep. The neural network approaches included (i)conventional single back-propagation neural networks (BPNNs), (ii) modular neural networks (MNNs) thatemploy a committee of several back-propagation neural networks, and (iii) quantum neural networks(QNNs). The QNNs were tested since some authors in other research areas have proposed that they arepotentially better at classification problems than conventional BPNNs, which can sometimes have difficultydistinguishing the boundaries between different classes. The neural networks were trained on combinationsof well logs using a genetically focussed methodology, which trains the networks on a short representativeinterval or genetic unit. This approach is potentially very effective in terms of cost and time. The lithologiespredicted by the various approaches were then compared with analysis of the cores for each well. Theresults for the well in the Nile delta showed that the QNNs overall outperformed the single BPNNs and theMNNs, and were particularly better at predicting the thin sand layers in the test intervals. The results for thewell in the Polish Carpathian Foredeep also showed that the QNNs were marginally the best at correctlyidentifying the sand intervals with depth in the simple Model 1, compared to the traditional statisticaltechniques (involving principal component analysis and discriminant analysis) and the other conventionalneural network approaches. QNNs also predicted the highest total number of correct lithologies with depthin the more detailed Model 2. In summary, this study indicated the potential of QNNs for improvinglithology classification in thin bedded gas turbidite reservoirs. The results also demonstrated that themethodology only requires a short representative interval to train the neural networks in order to delivergood predictions in the much larger test intervals.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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 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".