بهبود مدل سازی توسعه فضایی شهرها با تلفیق روشهای یادگیری ماشین و مدل CA Markov (مطالعه موردی کلان شهر قم).

Autor: صادق بولاقی, مسعود مینانی, حسین شفیع زاده مق, امید علی خوارزمی
Zdroj: Journal of Geography & Regional Development; Sep2023, Vol. 21 Issue 3, p199-232, 36p
Abstrakt: One of the inevitable consequences of the ever-increasing growth of the world’s population is the expansion of urbanization. So, it is very important to provide a vision of the spatial development of cities with the aim of understanding the correct pattern of city growth and providing the necessary infrastructure. Since Qom metropolis faced urban growth and has recorded 95% urbanization, this research focused on the spatial development of urban lands around this metropolis. First, the land use/cover and the urban growth merit maps were produced. Land use/ land cover maps of the region for the years 2000, 2010 and 2020 were produced using the random forest method in the Google Earth Engine, and the urban growth merit map for the years 2000 and 2010 was produced using MCDM analyses based on GIS. Finally, based on the ANN-CAMarkov and SVM-CA-Markov algorithms, 2020’s land use/cover maps were simulated. The validation of the models showed that the SVM-CA-Markov algorithm with the ROC (0.96) was more accurate and was chosen as the optimal algorithm for modeling the horizon of 2040. The results indicate the increasing spatial development of this metropolis. The area of urban land in this region will increase from 139.62 square kilometers in 2020 to more than 183 square kilometers in 2040. The evaluation of the results can help the relevant managers in order to take necessary policies to manage the situation as best as possible. This importance can be realized through planning for the regular development of the road network, the expansion of urban green spaces, etc. official organizations and local officials should have a purposeful monitoring of this issue, while having full control over the development directions of this metropolis. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index