A Classification Model of Urban Fire Level with Stacking Ensemble Learning

Smart firefighting is increasingly critical in enhancing fire safety and decision-making during fire incidents. A pivotal component of this intelligent approach is the accurate classify of fire levels, which equips fire departments with the insights needed to implement more effective countermeasures, thereby mitigating the damages inflicted by fires. In this paper, we collected and processed of urban fire data from diversified sources. Subsequently, we introduce a Stacking ensemble learning algorithm tailored for the classification of fire levels. Our experimental findings indicate that the model achieves a classification accuracy of 95.4% for fire levels. The research could significantly bolster the efficiency of urban fire emergency response strategies.