Carbon footprint forecasting using time series data mining methods: the case of Turkey

For a sustainable world, the biggest step to be taken in the fight against climate change is to reduce emissions from human activities. For this reason, it is important to work on reducing the carbon footprint, which is accepted as the measure of carbon emissions released to the world. In this direction, the study includes the estimation of Turkey's greenhouse gas emission rates for the next period. In the study, the greenhouse gas emission rates of Turkey were estimated using linear regression, multilayer perceptron and sequential minimal optimization (SMOreg) algorithms from time series data mining methods, utilizing the parameters of population, gross domestic product, energy production and energy consumption for the years 1990-2017. The greenhouse gas emission rate between 2018 and 2030 was estimated by using the SMOreg algorithm, which has the smallest error statistics and gives the closest values to the real values. As a result of the analysis, it has been found that as long as Turkey continues its current course, the amount of greenhouse gas emissions will increase gradually and reach 728,301 metric tons of CO2 equivalent in 2030. The estimated values reached for Turkey, which is included in most of the international conventions on climate change, are below the rate committed in the Paris climate convention. Although this situation is considered promising for Turkey, it is thought that policy makers should determine policies in this direction, without ignoring the estimates in other studies.