An Enhanced Support Vector Machine Model for Classification of Transit-Oriented Development

Transit-oriented development (TOD) can invigorate sustainable development by conveying a more coordinated transit and surrounding land use. Given the dynamic nature of urban community development, urban planners find it hard to precisely respond to questions such as where TOD planning around the transit hubs can succeed in the city. The present study proposes a framework utilizing a support vector machine (SVM) to enhance the TODness prediction of an area to address this issue. An SVM model has successfully applied to 16 bus rapid transit station areas in Bhopal city, India, using the tenfold cross-validation resampling methods and thirteen predictor variables. The models performance was in good agreement with 93.75% precision, utilizing the sigmoid kernel function and the regularization parameter esteem equivalent to 4. This methodology could be used at any scale, and the outcomes could offer recommendations for more accurate urban planning, fortifying the relationship between TOD and spatial association. The study provides the basis for predicting better future TODness classification, which will help the urban planner for sustainable urban planning and policy making.