A Machine-Insight-Based Geocomputational Approach to Decode the Complexities of the Restless Landscapes: Toward Integration of Imaginations and Planning Interventions

The proliferating urban influence brought a complex sociospatial, economic structure over the urban periphery. Previous studies focused on different approaches to capturing this intricate process; generalization was done by setting their own criteria. This study builds a decision-making model to comprehend the nature of Varanasi city’s periphery. Settlement-wise data are collected to calculate the study’s initial parameters: functional index, physical infrastructure dynamics score, physiological density, and nonagricultural land to total land ratio. The independent component analysis calculates the composite score. The learning cloud incorporates settlement-wise data in an encoded format to run the XGBoost machine learning algorithm. The findings highlight a competitive pattern of settlements and identify four periurban zones based on the deviation diffusion of calculated parameters. The study implements the Monte Carlo simulation for visualizing periurban decision space and recorded an asymptotic one-sample Kolmogorov–Smirnov test score of 0.051 at a <i>p</i> value of 0.076. The findings suggest four peripheral settlement clusters where the Gaussian mixture model confirms the reliability. This study highlights the importance of these clusters in attracting government and private players to invest and innovate functioning urban like institutions. This study’s methodological framework and inclusive approaches for city region planning and policy formulation remain significant.