Sustained efforts and investments in different sectors are essential for the overall development of a region. Various studies around the globe underscore the importance of investment in road transport infrastructure in many developing countries to achieve their development targets. The relationship between infrastructure investment and economic growth is often found to be inconsistent. This ambiguity leads to a lack of consensus on an appropriate scale of investment required among the policy makers. So, it is often necessary to depend on empirical evidence by developing causality direction, which significantly contributes to policy implications in developing countries. The objective of this analytical study is developing a relationship between the road transport infrastructure and economic growth of Nepal. For this, relevant data from 1998 to 2022 were used to perform a unit root test and determine the order of integration, followed by cointegration analysis to determine the long-run relationship between the variables. In addition, the vector error correction model (VECM) was employed to find the direction of causality. The findings indicate unidirectional long-run causality from gross capital formation, exports of goods and services, expenditure on road transport infrastructures, and road length to the GDP of Nepal. Furthermore, the expenditure on road transport infrastructures is observed to have a short-run impact on economic growth. This study recommends that a suitable transportation policy should be implemented to boost investment on road transport infrastructures to achieve sustainable economic growth in Nepal-like developing nations.

This study presents a comparative framework for evaluating the predictive performance of four machine learning models namely Gradient Boosting Machine (GBM), Random Decision Forest (RDF), Non-Parametric Regression (NP), and Decision Tree (TREE), in estimating the Unconfined Compressive Strength (UCS) of nano-doped fly ash reinforced clayey soil. The key innovation lies in combining ensemble learning with sensitivity, monotonicity, and SHAP (SHapley Additive exPlanations) analyses to enhance predictive accuracy and interpretability in geotechnical applications. Using a comprehensive dataset of key variables (Curing Days, Maximum Dry Density (MDD), Optimum Moisture Content (OMC), Fly Ash, Multi-Walled Carbon Nanotubes (MWCNT), and Sodium Hexametaphosphate (SHMP)) models were trained and validated using various statistical metrics (R², MAE, MSE etc.). GBM achieved the best performance (R²: 1.000, 0.955; MAE: 0.001, 0.022; MSE: 0.000, 0.001 in training and testing respectively), consistently outperforming RDF, NP, and TREE. Taylor diagrams, REC curves, and AOC analysis further confirmed GBM’s superior generalization and minimal error rates. Sensitivity analysis identified Days (0.4556), MDD (0.2458), and SHMP (0.1558) as the most influential factors, trends that were corroborated by SHAP analysis. Monotonicity analysis validated positive relationships with Days (R² = 0.9845) and MDD (R² = 0.689), while OMC and SHMP showed inverse effects. These results establish GBM as a highly accurate and interpretable model for UCS prediction, offering a powerful tool for optimizing soil stabilization in construction and geotechnical engineering.