An Intelligent Decision-Support Framework Using Linguistic Assessment and Probabilistic Neural Networks for Rapid Prioritization of Urban Building Stocks

OpenAlex English

Rapid and reliable tools are essential for urban regeneration and disaster risk mitigation in earthquake‐prone regions. This study proposes a hybrid decision‐support framework that integrates linguistic expert assessments with a probabilistic neural network (PNN) for rapid prioritization of urban building stocks. Eleven seismic vulnerability parameters were evaluated using a five‐level linguistic scale and applied to 194 buildings affected by the 2023 Kahramanmaraş earthquake in Türkiye. The model achieved 100% training accuracy and 95% testing accuracy, with misclassifications limited to adjacent damage classes. The proposed framework can support municipalities in rapidly identifying high‐risk buildings during urban regeneration planning. PCA‐based visualization confirmed clear class separation and consistent generalization. A post hoc sensitivity analysis further quantified decision‐boundary uncertainty, showing that near‐boundary transitions were concentrated in a small number of influential parameters. Class‐level stability indices indicated higher robustness for severe‐damage predictions. Overall, the proposed framework offers a practical and operationally efficient tool for first‐level seismic vulnerability screening and evidence‐based urban regeneration prioritization.