Abstract
<title>Abstract</title> <p>In many developing cities, paratransit is the layer of the transport system that absorbs the demand the formal network cannot serve, including during disruptions, which makes the dependability of this informal fleet a practical component of urban transport resilience. That dependability rests almost entirely on individual drivers, yet driver-level risk in the sector remains poorly characterized by conventional statistical approaches. This study develops and validates machine learning classification models to identify high-risk paratransit drivers in Gazipur, Bangladesh, using driver interviews (n = 507). Here employed five classification algorithms Logistic Regression, Random Forest, XGBoost, Gradient Boosting, and Stacking Ensemble to predict driver risk categories based on socioeconomic, behavioral, and operational characteristics. Results show that Logistic Regression achieved the best performance (F1-score = 0.483, AUC-ROC = 0.676), improving recall over traditional Poisson regression by 40% and F1-score by 12.6%. Feature importance analysis identified motorized three-wheeler operation as the dominant predictor of high-risk status, associated with roughly three times the crash risk of non-motorized modes, alongside daily income and driving experience. These findings point to mode-specific regulation of the informal paratransit fleet. The classification framework provides transportation authorities with an actionable tool for targeted safety interventions in informal transport systems.</p>