To improve the accuracy of lithology identification in uranium mine borehole strata and address the limitations of traditional ensemble learning models in achieving satisfactory lithology recognition performance, this study proposes a PCA-optimized Stacking ensemble learning model. The methodology begins by quantitatively evaluating the linear correlation strength between various logging parameters and the target lithology using Pearson correlation coefficients. This statistical analysis is further refined by incorporating geophysical mechanisms of well logging, leading to the selection of six logging parameters that demonstrate strong lithological relevance as input features. These parameters effectively capture essential petrophysical characteristics while minimizing redundancy in the feature set. Simultaneously, the model employs a dual-faceted approach to base learner selection. The Pearson correlation coefficients of prediction errors among base learners are systematically calculated to assess their error interdependence. Additionally, the Q-statistic matrix is utilized to evaluate the correlation and diversity of prediction outcomes. From this, a combination of base models with strong error complementarity (i.e., low correlation) and significant prediction mode differences (i.e., low Q-value) is selected. To optimally consolidate the predictive information, the PCA algorithm is utilized to perform a weighted fusion of the prediction results derived from the diverse base models. The resulting principal components, which represent the most discriminative fused features, then form the refined input feature space for training the second-layer meta-model, thereby achieving a sophisticated and high-accuracy multi-level ensemble learning framework. Experimental results show that the PCA-optimized Stacking model achieves a recognition accuracy of 97.19%, significantly outperforming traditional Stacking models and all individual models. The experimental results indicate that the recognition accuracy of the PCA-optimized Stacking model reaches 97.19%, significantly outperforming both the traditional Stacking model and all individual models. Furthermore, the study provides valuable insights and a novel technical framework for lithology identification in sandstone-type uranium mine borehole strata, contributing to more reliable and efficient geological interpretation in uranium exploration and mining operations.