Corrigendum to “Prediction of oil rates using Machine Learning for high gas-oil ratio and water cut reservoirs” [Flow Meas. Instrum. 82 (2021) 102065] (Flow Measurement and Instrumentation (2021) 82, (S0955598621001680), (10.1016/j.flowmeasinst.2021.102065))

The authors regret The paper presents the application of machine learning to generate different models predicting the production rates for high Gas-Oil Ratio (GOR) and high Water Cut (WC) wells. It is an extension of our earlier results reported in Ibrahim et al., 2021 [1]. The current study presents the results obtained applying adaptive network-based fuzzy inference systems (ANFIS), and functional networks (FN), while [1] shows the application of support vector machines (SVM) and random forests (RF). These techniques provide different learning processes applied to a common dataset and generate a range of model performances. SVM is a supervised learning method with an associated learning algorithm that analyzes data and recognizes patterns of input/output data. The SVM technique builds input prototypes in a space with higher dimensionality by employing a nonlinear mapping method. The RF model is made up of decision trees that use the ensemble learning method for regression. In the current study, ANFIS combines the concepts of fuzzy logic and neural networks to form a hybrid intelligent system that enhances the ability to automatically learn and adapt. FN uses multi-argument functional models associated with each neuron for processing and learning from the data obtained. The common issue associated with SVM and RF is overfitting, which is demonstrated especially in the RF results in Ref. [1], where the model accuracy was high during the training process (R² = 0.98) but was significantly reduced for the testing data set. In the current work, the overfitting issue is addressed, particularly for the ANFIS model, with R² = 0.95 for both training and testing datasets. The authors would like to apologise for any inconvenience caused.