Review of: "Uncertainty-Driven Modeling of Microporosity and Permeability in Clastic Reservoirs Using Random Forest"

The manuscript introduces a robust and innovative approach for predicting permeability and microporosity in clastic reservoirs using Random Forest, enhanced with uncertainty-driven resampling.The integration of accessible geological parameters and the aim to reduce reliance on costly laboratory techniques are commendable.However, several conceptual and methodological issues must be clari ed to improve the manuscript's clarity, scienti c rigor, and reproducibility.Below are major issues to address: 1.The manuscript includes the Th/K (thorium to potassium) ratio as one of the input variables but does not explain its relevance to microporosity or permeability prediction.Clarify the geochemical or petrophysical rationale for including this ratio.Is it used as a proxy for clay type or maturity?If so, support this with references and explain how it relates to uid ow characteristics or diagenetic trends in your study area.2. A signi cant methodological gap exists in the description of how the dataset expanded from 41 samples to 61,600 data points.• Was this based on Monte Carlo simulation, bootstrapping, perturbation with Gaussian noise, or another resampling strategy?• How was data correlation preserved during this augmentation?• Were categorical variables (e.g., permeability class) held constant during sampling or also reclassi ed? Include this information in the methodology section and consider illustrating the resampling pipeline with a schematic.3 Despite the large synthetic dataset, the original sample size (41) is limited.Acknowledge this limitation more explicitly and discuss whether the model is potentially over tting to noise within a small Qeios qeios.com