پیوندهای مفید
اخبار و اعلانات
آمار
| تعداد نشریات | 163 |
| تعداد شماره‌ها | 6,878 |
| تعداد مقالات | 74,135 |
| تعداد مشاهده مقاله | 137,876,120 |
| تعداد دریافت فایل اصل مقاله | 107,235,226 |
Predictive Analytics for Urban Sprawl Using Machine Learning in Land Cover Mapping | ||
| Earth Observation and Geomatics Engineering | ||
| مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 15 شهریور 1404 | ||
| نوع مقاله: Original Article | ||
| شناسه دیجیتال (DOI): 10.22059/eoge.2025.393376.1173 | ||
| نویسندگان | ||
| Mohsen Niroomand؛ Parham Pahlavani* | ||
| School of Surveying and Geospatial Engineering, College of Engineering, University of Tehran, Tehran, Iran. | ||
| چکیده | ||
| Urban sprawl, characterized by low-density, uncoordinated, and outward urban expansion, presents critical challenges to sustainable development, particularly in rapidly growing metropolitan regions such as Tehran. This study employs an integrated framework combining remote sensing, spatial urban sprawl indicators, and advanced machine learning techniques to analyse and project land cover changes between 2011 and 2026. Initial land cover maps for the years 2011, 2016, and 2021 were generated using the Random Forest (RF) algorithm applied to Landsat 7 and 8 imagery, achieving overall classification accuracies of 92.53%, 93.27%, and 93.88%, respectively. Subsequently, a comprehensive set of urban sprawl indices—derived from census data, transportation networks, and land use parcel information—was utilized alongside land cover transition maps to train Multi-Layer Perceptron (MLP), Decision Forest (DF), and Support Vector Machine (SVM) models within a Markov chain framework. Dimensionality reduction techniques, including Principal Component Analysis (PCA) and Independent Component Analysis (ICA), were applied to enhance model efficiency. Among the evaluated models, the MLP trained with the complete feature set demonstrated superior performance, attaining an F1-score of 83.95%. The projections suggest a 6% increase in built-up areas by 2026, predominantly at the expense of barren lands and green spaces. The results underscore the potential of integrating geospatial technologies with machine learning methodologies to support data-driven urban planning and the formulation of sustainable land management policies in rapidly urbanizing contexts. | ||
| کلیدواژه‌ها | ||
| Urban Sprawl؛ Land Cover Change Modelling؛ Machine Learning Algorithms؛ dimensionality reduction؛ Markov Chains | ||
|
آمار تعداد مشاهده مقاله: 27 |
||
سامانه مدیریت نشریات علمی. قدرت گرفته از سیناوب