Adaptive Cluster-Based Normalization for Robust TOPSIS in Multicriteria Decision-Making

In multicriteria decision-making (MCDM), methods like TOPSIS play a critical role in evaluating and comparing alternatives across multiple criteria. However, traditional normalization techniques often fall short when faced with data that includes outliers, large variances, or diverse measurement units, leading to rankings that may be skewed or biased. To address these challenges, this paper introduces an adaptive cluster-based normalization approach, applied to a real-world case study of selecting a city to host an international event. The method works by grouping alternatives into clusters based on their similarities in criterion values and applying tailored normalization within each group. This localized approach reduces the impact of outliers and ensures that scaling adjustments are aligned with the unique characteristics of each cluster. When tested in the case study, where cities were evaluated based on cost, infrastructure, safety, and accessibility, the cluster-based normalization method provided more stable and balanced rankings, even in the presence of significant data variability. By improving fairness and adaptability, this approach strengthens TOPSIS’s ability to deliver accurate, context-aware decisions, making it an invaluable tool for tackling complex datasets in real-world applications.