<![CDATA[Accurately identifying lithological facies is crucial for comprehending geological variations in a proven reservoir. To enhance the accuracy of facies classification compared to previous studies on the same dataset, five distinct machine learning algorithms were employed to predict facies in both a panoma field dataset and Z-Field, Indonesia. The analysis data samples with known facies, originating from core data from Panoma Field and Z-Field. Facies classification was addressed using five well-known classification algorithms, namely K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Neural Network Classifier (NNC), Random Forest Classifier (RFC), and Decision Tree Classifier (DTC). The dataset was divided into training and testing subsets to evaluate the machine learning models. The five suggested algorithms demonstrate effective facies prediction, closely aligning with the actual facies in the test wells within the Panoma field. However, these algorithms struggle to predict facies accurately in the Z field well, primarily attributed to the imbalanced data distribution between sandstone-claystone and siltstone-limestone. Equalizing the number of facies labels in the training data becomes essential to enable the algorithm to recognize patterns and accurately estimate all facies types]]>
Copyrights © 2024
