Understanding and modeling land use and land cover changes are critically important subjects for the purposes of environmental management, planning, and civil engineering. Models are considered as good approaches for having better estimation or understanding of the phenomenon of land use, land cover changes, thoughtful planning, and alteration in better management of cities and villages. Therefore, this study was conducted to analyze and compare the three models of logistic regression, ANNs, and Geo-Mod in predicting the expansion of built-up areas in a 10-year period in Marivan city, Iran. For this purpose, land cover maps at different times were prepared using Landsat data, including the images of the TM sensor of Landsat 5 in 1989, the ETM+ sensor of Landsat 7 in 2000, and the TM sensor of Landsat 5 in 2011. According to the results, during the period from 1989 to 2011, the built-up area of land cover has been multiplied by 2.74 (73.63 ha/year), which is equal to 1296 ha in 2011 compared to 459 ha in 1989. According to the results, logistic regression has been the preferred method in modeling changes to built-up areas among the presented methods. In addition, one of the advantages of logistic regression is to determine the relationship between variables and changes to the built-up areas. Although Geo-Mod yielded poorer results than the other two methods, its advantages of the neighborhood rule and the maximum use of available data should not be disregarded.

中文翻译：

---

使用逻辑回归、人工神经网络和 Geo-Mod 模型模拟伊朗马里万市建成区的扩张

理解和模拟土地利用和土地覆盖变化对于环境管理、规划和土木工程而言是至关重要的主题。模型被认为是更好地估计或理解土地利用现象、土地覆盖变化、周密规划和改变以更好地管理城市和村庄的好方法。因此，本研究旨在分析和比较逻辑回归、ANNs 和 Geo-Mod 三种模型在预测伊朗 Marivan 市 10 年建成区扩张方面的作用。为此，利用 Landsat 数据制作了不同时期的土地覆盖图，包括 1989 年 Landsat 5 的 TM 传感器图像、2000 年 Landsat 7 的 ETM+传感器和 2011 年 Landsat 5 的 TM 传感器图像。结果，1989 年至 2011 年期间，土地覆盖建成面积乘以 2.74（73.63 公顷/年），相当于 2011 年的 1296 公顷，而 1989 年为 459 公顷。根据结果，逻辑回归在所提出的方法中，已成为对建筑区域的变化进行建模的首选方法。此外，逻辑回归的优势之一是确定变量与建成区变化之间的关系。尽管 Geo-Mod 的结果比其他两种方法差，但不应忽视其邻域规则和最大利用可用数据的优势。在所提出的方法中，逻辑回归一直是对建成区变化进行建模的首选方法。此外，逻辑回归的优势之一是确定变量与建成区变化之间的关系。尽管 Geo-Mod 的结果比其他两种方法差，但不应忽视其邻域规则和最大利用可用数据的优势。在所提出的方法中，逻辑回归一直是对建成区变化进行建模的首选方法。此外，逻辑回归的优势之一是确定变量与建成区变化之间的关系。尽管 Geo-Mod 的结果比其他两种方法差，但不应忽视其邻域规则和最大利用可用数据的优势。