Hybrid integration of TOPSIS and machine learning models in a scenario-aware framework for shelter prioritization under conflict conditions

Determining shelter sites in conflict-affected areas poses a major challenge for humanitarian agencies, faced with time limitations, limited data, and uncertainty. This study presents a Hybrid Scenario-Aware Decision Support Framework that integrates Multi-Criteria Decision-Making (MCDM) and Machine Learning (ML), offering transparent, scalable, and data-driven shelter evaluation. At first, the TOPSIS technique was used to rank shelter sites based on indicators of accessibility, occupancy capaci