Missing intervals or unreliable well-log measurements pose a persistent challenge for subsurface characterization, particularly for curves critical to porosity, lithology, and velocity analysis. This study investigates two machine-learning strategies for reconstructing missing logs: a hybrid Multivariate Imputation by Chained Equations with Gradient Boosted Trees (MICE+GBT), and also Convolutional Neural Network (CNN). The prediction targets are three essential logs-neutron porosity (NPHI), bulk density (RHOB), and compressional travel time (DT). Each reconstructed artificial log from complementary measurements including resistivity, Photoelectric Factor (PEF), and spectral gamma-ray. A rigorous preprocessing workflow was applied, followed by evaluation under a well-level cross-validation scheme to simulate deployment on unseen wells. Performance was measured using Mean Squared Error (MSE), Mean Absolute Error (MAE), correlation coefficient (R), and coefficient of determination (R²).  This research utilizes two separate methods to estimate absent well logs in the Sarvak Formation, a diverse carbonate reservoir noted for intricate pore configurations and variability in lithology. These approaches tackle deficiencies in log data while assessing performance in regulated environments. In the following section, we will outline the dataset, prediction objectives, preprocessing steps, workflows, and evaluation methods. Results indicate that the MICE+GBT approach consistently outperforms Convolutional LSTM across all three target logs, particularly for intervals affected by washouts or tool measurement failures. Beyond statistical performance, the reconstructed curves preserved geological consistency in density–neutron and sonic–density trends, ensuring reliability for downstream reservoir interpretation. The findings demonstrate the practical benefits of adapting ensemble-based imputation methods to petrophysical data, providing a robust and interpretable framework for improving log data alliance in reservoir studies. Results demonstrated that the MICE+GBT workflow consistently delivered accurate, stable, and interpretable predictions across both Danan and Azadegan fields. Its strength was most evident in DT reconstruction, where ConvLSTM struggled with convergence and produced elevated error magnitudes. Ensemble models also yielded narrower error distributions and superior generalization across wells, underscoring their robustness in data-limited carbonate environments. Conversely, the ConvLSTM framework captured local depth-wise dependencies effectively, particularly for NPHI, and showed potential when sufficient training data were available. However, its sensitivity to heterogeneity and tendency toward error dispersion limit its reliability in reservoirs with strong lithological variability. Overall, the findings suggest that ensemble methods presently offer more dependable solutions for carbonate reservoirs with restricted datasets, while deep learning approaches hold promise for future applications given larger, more diverse training corpora. Integrating the interpretability and stability of ensemble models with the representational power of deep learning could form the basis of next-generation workflows for missing log prediction.