EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Hybrid Spatial Multi-Criteria Decision-Making Framework Based on Information Divergence Optimization

Xiaoyun Pei, LS Lee

Contemporary Mathematics · 30.06.2026 · 4048-4064

DOI: 10.37256/cm.74202610158 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Contemporary Mathematics, online erschienen am 30.06.2026. Verfasst von Xiaoyun Pei und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Multi-criteria decision-making (MCDM) over heterogeneous spatial data presents two fundamental challenges that remain insufficiently addressed in the literature: the optimal fusion of subjective and objective weighting schemes, and the explicit incorporation of spatial dependence into distance-based ranking models. This paper proposes a hybrid spatial MCDM framework that addresses both issues through rigorous mathematical construction. First, we formulate combined weighting as a minimum Kullback-Leibler (KL) divergence optimization problem and derive a closed-form solution via the Lagrange multiplier method, yielding a weighted geometric mean combination that optimally balances expert judgment against data-driven information. Second, we introduce a spatial-weighted TOPSIS model that explicitly integrates a spatial weight matrix into the Euclidean distance computation to the positive and negative ideal solutions, thereby encoding Tobler's first law of geography into the ranking mechanism. Third, we establish several theoretical properties of the framework: a sensitivity theorem bounding the error propagation under data perturbation, a robustness theorem for the KL-divergence combined weights, and a rank-reversal analysis proving that a globally bounded normalization scheme eliminates the classical TOPSIS rank-reversal pathology. The framework's practical utility is demonstrated through a large-scale numerical experiment involving 2,502 spatial alternatives with 10 heterogeneous attributes derived from multi-source geographic data. Comparative analysis against standard AHP-TOPSIS and entropy-TOPSIS baselines confirms that the proposed framework produces spatially coherent, numerically stable, and theoretically grounded priority rankings. The framework is general and transferable to any spatial MCDM problem characterized by heterogeneous attribute data and spatial dependence.

Zitieren

APA

Pei, X., & Lee, L. (2026). A Hybrid Spatial Multi-Criteria Decision-Making Framework Based on Information Divergence Optimization. Contemporary Mathematics, 4048-4064. https://doi.org/10.37256/cm.74202610158

Vancouver

Pei X, Lee L. A Hybrid Spatial Multi-Criteria Decision-Making Framework Based on Information Divergence Optimization. Contemporary Mathematics. 2026:4048-4064. doi:10.37256/cm.74202610158

BibTeX
@article{Pei2026,
  author = {Pei, Xiaoyun and Lee, LS},
  title = {A Hybrid Spatial Multi-Criteria Decision-Making Framework Based on Information Divergence Optimization},
  journal = {Contemporary Mathematics},
  year = {2026},
  pages = {4048--4064},
  doi = {10.37256/cm.74202610158}
}
Kostenlos starten →