Machine Learning Techniques in Enhanced Oil Recovery Screening Using Semisupervised Label Propagation
Bibliographic record
Abstract
Summary Efficiently choosing the optimal enhanced oil recovery (EOR) technique is a critical requirement in reservoir engineering. Machine learning (ML) methods, with a well-established history of application, serve as a swift and dependable tool for EOR screening. In this paper, we aim to evaluate the effectiveness of various ML algorithms for EOR screening, utilizing a comprehensive database of nearly 1,000 EOR projects. This study delves into a comprehensive evaluation of regression and classification-based algorithms to develop a reliable screening system for EOR predictions and address challenges such as limited labeled data and missing values. Our analysis considered various EOR processes, including gas injection, chemical, and thermal EOR techniques. Various ML methods such as random forest (RF), K-nearest neighbors (KNN), support vector machine (SVM), shallow artificial neural networks (SANN), naive Bayes classifier (NBC), logistic regression (LR), and decision tree (DT) are applied, enabling both intermethod comparisons and evaluations against advanced methods, multiobjective deep artificial neural networks (MDANN), and multiobjective artificial neural networks (MANN). These advanced techniques provide the unique capability to concurrently address both regression and classification tasks. Considering that conventional methods can only be implemented on a single task, the RF, MANN, MDANN, and KNN algorithms demonstrated top-tier performance in our classification analysis. Regarding the regression task, KNN, RF, and MDANN displayed exceptional performance, signifying their prowess in predictive accuracy. However, MANN exhibited moderate performance in regression analysis. In addition, our study identified areas where certain algorithms, such as support vector regression (SVR), exhibited weaker performance, highlighting the importance of comprehensive model evaluation. This paper contributes novel insights into the application of ML techniques for EOR screening in the petroleum industry. By addressing challenges such as limited labeled data and missing values and by providing a thorough evaluation of various ML algorithms, our study offers valuable information for decision-makers in the oil and gas sector, aiding in the selection of suitable algorithms for EOR projects. In addition, the use of semisupervised label propagation and advanced techniques like KNN imputation adds to the existing body of literature, enhancing the state of knowledge in this domain.
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 | 0.004 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".