Comprehensive review on predicting CBR values using machine learning techniques

Evaluating the subgrade bearing capacity using the California bearing ratio test is necessary in infrastructure projects. The California Bearing Ratio (CBR) is a critical parameter in geotechnical engineering, particularly in the design of pavements and subgrade materials. Traditional methods for predicting CBR, such as empirical correlations and laboratory tests, are often time-consuming, labor-intensive, and limited in capturing complex interactions between soil properties and external factors. Machine learning (ML) has emerged as a powerful tool for addressing these limitations, offering the potential to predict CBR with greater accuracy and efficiency. This review paper aims to provide a comprehensive overview of the application of machine learning techniques for CBR prediction. The methodology involves a systematic review of existing literature, focusing on studies that employ ML models such as Artificial Neural Networks (ANN), Support Vector Machines (SVM), and Random Forest (RF). Key findings from the reviewed studies are summarized, highlighting these techniques, the performance metrics, and the dataset size. The paper also discusses the advantages and limitations of ML in CBR prediction, including challenges related to data quality, model interpretability, and generalizability.