prog-co2
Forecast for 3 methods of US emissions of CO2 to the atmosphere
Currently, the United States is the second country that concentrates the greatest amount of CO2 emissions into the atmosphere. In order to improve future environmental policies, this series of time is analyzed with the data on CO2 emissions. Within the present work three models of forecasts of C02 emissions are used. In the first instance by a linear method, Linear Autoregression. Then two non-linear models are used, on the one hand, LS-SVM with Gaussian kernel. On the other hand with an ANN with Levenberg Marquardt learning algorithm.