This study investigates the potential of nanoclay as an effective stabilizing agent for expansive clayey soils and evaluates the performance of machine learning (ML) models in predicting key geotechnical properties. Laboratory experiments were conducted by varying nanoclay content to assess its influence on Free Swell Index (FSI) and Maximum Dry Density (MDD). The results showed a significant reduction in FSI from 29.21% in untreated soil to 8.99% with 3% nanoclay indicating improved volumetric stability. MDD increased to 16.51 kN/m3 at 3% nanoclay due to enhanced particle packing, although higher dosages led to a decline in density due to excess moisture retention. To complement the experimental work, regression and ML models, including Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Regression (SVR), were developed to predict FSI and MDD. Among these, RF demonstrated superior performance during training (R2 = 0.999, MSE = 0.72), while KNN showed the best generalization during testing (R2 = 0.996). SVR lagged in accuracy across all phases. The findings affirm that nanoclay is a promising soil stabilizer and that ML models, particularly RF and KNN, can effectively predict geotechnical responses, reducing the need for extensive physical testing.