Smart Urban City Building Criteria Based on Carbon Footprint of Individuals Using Deep Neural Network

Abstract

The definition of an urban city is a highly populated place that also has a large concentration of non-farming jobs, buildings, and infrastructure.  Structures such as homes, businesses, roads, and transportation networks make up its built environment.  Rural regions are often characterized by lower population densities and an increased dependence on agriculture, in contrast to urban areas.  There are more people per unit of land area in urban areas than in rural ones.  In addition, man-made structures and vast infrastructure characterize urban areas.  Because of this, many city dwellers work in professions unrelated to rural areas. Thus, to design an urbanism city, one should consider the carbon footprint. Various works were achieved to predict the carbon footprint emission. In this work, a deep neural network (DNN) was suggested to predict the carbon footprint in urban cities. The proposed DNN consisting mainly of three dense layers, with other layers such as activation layers and normalization. Results show that the suggested model performs well in predicting carbon footprint with R²-score equals to 0.5. The utilized dataset in this work is available online at Kaggle.com, and it is well balanced, therefore, there will be no need to do balancing operations. Hence, the trained model can be utilized in urban cities as a robust design tool.