Identifying underutilized land by eXplainable artificial intelligence and geographic similarity ensemble model with limited samples

IF 7 1区经济学 Q1 DEVELOPMENT STUDIES

Habitat International Pub Date : 2025-07-08 DOI:10.1016/j.habitatint.2025.103503

Chuan Lin , Yongfang Huang , Yilun Liu , Guang Li , Zegen Zhou , Yuanjun Zhong , Hongmei Wang , Jinggang Li

{"title":"Identifying underutilized land by eXplainable artificial intelligence and geographic similarity ensemble model with limited samples","authors":"Chuan Lin , Yongfang Huang , Yilun Liu , Guang Li , Zegen Zhou , Yuanjun Zhong , Hongmei Wang , Jinggang Li","doi":"10.1016/j.habitatint.2025.103503","DOIUrl":null,"url":null,"abstract":"<div><div>As cities globally confront the dual challenges of spatial resource scarcity and aging urban fabrics, the precise identification of underutilized land emerges as a critical pathway toward sustainable urban regeneration. However, persistent methodological gaps hinder precise identification due to three unresolved scientific problems: (1) multifactorial spatial complexity obscuring determinant interactions, (2) limited sample availability constraining machine learning efficacy, and (3) opaque decision-making processes in conventional algorithms. This study resolves these through an eXplainable Artificial Intelligence-Geographic Similarity Reasoning (XAI-GSR) model integrating three innovations: a multidimensional indicator system quantifying land-use efficiency across morphology, economic, social, and ecological dimensions; XGBoost-SHAP interpretation elucidating nonlinear factor contributions; and geospatial analogical reasoning overcoming sample scarcity. Applied to Shenzhen, the model achieved 82.9 % accuracy, identifying 9668 underutilized blocks (25.44 % total) with distinct typological distribution - Type 1 (6.99 %) reflecting central district efficiency versus Type 2 dominance (56.17 %) revealing suburban improvement potential, while Type 3 (27.18 %) and Mixed-type (9.67 %) clusters predominantly occupy eastern/northern low-density zones. Compared to existing methods, our framework advances underutilized land detection by simultaneously resolving sample limitations through geospatial similarity reasoning and enhancing reliability via uncertainty-quantified similarity metrics, providing urban planners with an empirically validated decision-support tool for targeted regeneration strategies.</div></div>","PeriodicalId":48376,"journal":{"name":"Habitat International","volume":"163 ","pages":"Article 103503"},"PeriodicalIF":7.0000,"publicationDate":"2025-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Habitat International","FirstCategoryId":"96","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S019739752500219X","RegionNum":1,"RegionCategory":"经济学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"DEVELOPMENT STUDIES","Score":null,"Total":0}

引用次数: 0

Abstract

As cities globally confront the dual challenges of spatial resource scarcity and aging urban fabrics, the precise identification of underutilized land emerges as a critical pathway toward sustainable urban regeneration. However, persistent methodological gaps hinder precise identification due to three unresolved scientific problems: (1) multifactorial spatial complexity obscuring determinant interactions, (2) limited sample availability constraining machine learning efficacy, and (3) opaque decision-making processes in conventional algorithms. This study resolves these through an eXplainable Artificial Intelligence-Geographic Similarity Reasoning (XAI-GSR) model integrating three innovations: a multidimensional indicator system quantifying land-use efficiency across morphology, economic, social, and ecological dimensions; XGBoost-SHAP interpretation elucidating nonlinear factor contributions; and geospatial analogical reasoning overcoming sample scarcity. Applied to Shenzhen, the model achieved 82.9 % accuracy, identifying 9668 underutilized blocks (25.44 % total) with distinct typological distribution - Type 1 (6.99 %) reflecting central district efficiency versus Type 2 dominance (56.17 %) revealing suburban improvement potential, while Type 3 (27.18 %) and Mixed-type (9.67 %) clusters predominantly occupy eastern/northern low-density zones. Compared to existing methods, our framework advances underutilized land detection by simultaneously resolving sample limitations through geospatial similarity reasoning and enhancing reliability via uncertainty-quantified similarity metrics, providing urban planners with an empirically validated decision-support tool for targeted regeneration strategies.

查看原文本刊更多论文

利用有限样本可解释人工智能和地理相似性集合模型识别未充分利用土地

随着全球城市面临空间资源稀缺和城市结构老化的双重挑战，精确识别未充分利用的土地成为实现可持续城市更新的关键途径。然而，由于三个尚未解决的科学问题，持续存在的方法差距阻碍了精确识别：(1)多因子空间复杂性模糊了决定因素的相互作用，(2)有限的样本可用性限制了机器学习的有效性，以及(3)传统算法中不透明的决策过程。本研究通过一个可解释的人工智能-地理相似推理（XAI-GSR）模型解决了这些问题，该模型集成了三个创新：一个多维指标体系，在形态、经济、社会和生态维度上量化土地利用效率；XGBoost-SHAP解释阐明非线性因素贡献；地理空间类比推理克服了样本稀缺性。应用于深圳，该模型的准确率达到82.9%，识别出9668个未充分利用的街区（总数25.44%），它们具有不同的类型分布——1型（6.99%）反映中心区域效率，2型（56.17%）显示郊区改善潜力，而3型（27.18%）和混合型（9.67%）集群主要占据东部/北部低密度区。与现有方法相比，我们的框架通过地理空间相似性推理解决了样本限制，并通过不确定性量化相似性度量提高了可靠性，从而推进了未充分利用的土地检测，为城市规划者提供了一种经验验证的决策支持工具，用于有针对性的再生策略。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Habitat International Multiple-

CiteScore

10.50

自引率

10.30%

发文量

151

审稿时长

38 days

期刊介绍： Habitat International is dedicated to the study of urban and rural human settlements: their planning, design, production and management. Its main focus is on urbanisation in its broadest sense in the developing world. However, increasingly the interrelationships and linkages between cities and towns in the developing and developed worlds are becoming apparent and solutions to the problems that result are urgently required. The economic, social, technological and political systems of the world are intertwined and changes in one region almost always affect other regions.