Floods are a significant natural disaster, inundating riverbanks and causing extensive damage. Particularly vulnerable to severe flooding during the monsoon season is the Subarnarekha River basin in eastern India, posing risks to communities, agriculture, and infrastructure. While preventing floods is challenging, technological advancements, such as machine learning in geospatial analysis, offer promising possibilities for identifying and managing flood-prone areas. This study utilizes machine learning boosting algorithms and 15 conditioning factors, including elevation, rainfall, and drainage density, to assess flood susceptibility in the Subarnarekha River basin. Historical flood data spanning 25 years (1998–2022) are employed to train and validate the models. Evaluation metrics such as precision, recall, F1 score, and area under the curve (AUC) demonstrate the effectiveness of the models, with AUC values ranging from 0.91 to 0.95. AdaBoost emerges as the most effective model, achieving an AUC of 95%, followed by XGBoost (93%), Gradient Boosting (92%), CatBoost (92%), and Stochastic Gradient Boosting (91%). The susceptibility analysis reveals varying flood hazard conditions across different regions, with upper reaches experiencing low hazard conditions and coastal areas facing high susceptibility due to heavy rainfall and runoff. This study emphasizes the utility of machine learning techniques in enhancing flood risk assessment and management strategies.