<p>One of the most important characteristics that determines the permeability of porous media is hydraulic conductivity, which has an impact on a number of technical and environmental processes. In this study, the influence of <i>Bacillus subtilis</i> on the horizontal permeability of stratified porous media is investigated at various orientations. Additionally, machine learning methods are used to predict the horizontal permeability coefficient. Experiments show that the presence of bacteria results in a significant reduction in hydraulic conductivity. This reduction is a consequence of the formation of biofilm and the blocking of pores. Higher bacterial densities are associated with greater reductions. Permeability increases with increasing orientation angle, potentially due to different flow pathways. Yet, bacterial obstruction remains the dominant factor. To simulate these permeability changes, several machine learning algorithms, including support vector machines (SVMs), artificial neural networks (ANNs), and gene expression programming (GEP), were evaluated. According to a statistical study, the GEP model achieved the highest prediction accuracy, outperforming other models on metrics such as accuracy. On the other hand, the SVM performed considerably worse than the ANN, which demonstrated strong predictive capabilities. The findings highlight the power of microbial interactions to influence the flow of subterranean water and highlight the effectiveness of machine learning techniques in predicting permeability.</p>