Modeling Urban Land Transformation In Malang City: A Cellular Automata Model With Artificial Neural Networks And Logistic Regression

This study aims to model urban land transformation in Malang City using a Cellular Automata (CA) approach integrated with Artificial Neural Networks (ANN) and Logistic Regression (LR). The model was developed to predict land-use changes over the next 10 years (2024-2034) by utilizing spatial data from 2014 and 2024. The method involves spatial analysis using Quantum GIS (QGIS) software with the MOLUSCE plugin, which enables the simulation of land cover changes based on transition probability matrices. The results show that the CA-LR model provides higher accuracy compared to the CA-ANN model, with a Kappa value reaching 1 at the location level. The simulations indicate a significant decrease in non-built-up land, from 4,090.85 ha in 2024 to 3,731.40 ha in 2044, while built-up land increased from 7,030.70 ha to 7,390.15 ha over the same period. Factors such as population growth, accessibility, and land prices were identified as the main drivers of land-use change. The findings of this study can serve as a reference for stakeholders in planning sustainable urban development, particularly in managing settlement growth and maintaining a balance between built-up areas and green open spaces.