Maximum Dry Density (MDD) and Optimum Moisture Content (OMC) are two important parameters of soil filling, which affect the soil stability and bearing capacity, and thus the reliability and durability of facilities such as highways and bridges. Therefore, it is important to make reasonable predictions of OMC and MDD. Four machine learning algorithms, namely, Support Vector Machine (SVM), Artificial Neural Network (ANN), Random Forest (RF), and Extreme Gradient Boosting Tree (XGBoost), are adopted in this paper to establish MDD and OMC prediction models. After training and testing, the best models of the four algorithms are compared. The results show that, as an ensemble learning algorithm, XGBoost is the best model for predicting MDD and OMC, with an R² of 0.9234 for OMC, and an R² of 0.9098 for MDD. Finally, the feature importance analysis concludes that the plastic limit (PL) and the liquid limit (LL) are the two features that affect OMC and MDD the most. The prediction of soil compaction parameters using machine learning models, especially ensemble learning, can significantly reduce the amount of laboratory work and improve the efficiency of optimizing design for soil resource utilization in engineering construction.

To investigate the effect of Polyvinyl Alcohol (PVA) fiber incorporation on the erosion resistance of soil under different environmental conditions, this study selected fiber length and fiber content as variables and systematically compared the performance difference between fiber-reinforced soil and plain soil through dry and wet cycles, freeze-thaw cycles, sulfate erosion, and scouring tests. The fiber soil was found to have significant erosion resistance compared to the plain soil. The results showed that, for dry-wet cycling, the addition of PVA fibers effectively enhances the soil sample's resistance. Particularly, when the fiber content ranges from 0.5% to 0.8%, the soil sample's resistance to dry-wet cycles is significantly improved. Moreover, the influence of fiber content on enhancing the dry-wet cycling performance of the sample is more pronounced than that of fiber length. Shrinkage tests showed that the soil undergoes three distinct stages during the drying process: constant rate, deceleration, and residual phases. Under the action of freeze-thaw cycles, specimens containing PVA fibers demonstrate good deformation resistance and high unconfined compressive strength and do not appear to be damaged in the process of freezing and thawing, indicating good freeze-thaw resistance. Meanwhile, the effect of fiber length on the strength-loss rate is minor, suggesting that the freeze-thaw resistance of the specimens is relatively insensitive to fiber length. For the sulfate erosion resistance test, with an increase in fiber content, the strength of the specimen improves, and its strength loss rate is significantly reduced, whereas the influence of fiber length on the specimen remains limited. In the scouring test, soil samples show good scour resistance after fiber content, and the scour growth rate decreases with time.