Ekolojik ayak izine çok etki eden faktörlerin yapay sinir ağı sınıflama özelliği kullanılarak belirlenmesi: Türkiye örneği

Abstract

Since the end of the 20th century, ecological problems have become a priority problem due to industrialization, urbanization, technological developments and rapid population growth. The change in human living standards causes many ecological problems such as unconscious consumption of natural resources, extinction of forests and living species. Ecological Footprint is developed to measure the demand pressure that people exert on the environment. In study, Neural Network Fitting Model was used in MATLAB, for the development Artificial Neural Network (ANN) by using the data of 1996-2018 to estimate Turkey's ecological footprint. Urban Population, Renewable Energy Consumption, R&D Expenditures and Human Development Index were chosen as independent variables. The data were obtained from the database of “World Bank Group” and “Human Development Reports”. For the ANN, Levenberg-Marquardt algorithm was used to determine the appropriate hidden layer and hidden neurons in each layer. The data used to train an artificial neural network using feedforward and backpropagation were randomly divided into three groups for training, testing and validation purposes. R values for each stage, respectively; 0.999, 0.948, was obtained as 1. According to the results obtained, the independent variable with the greatest effect on the ecological footprint was found to be the Urban Population.