Adaptive-Neuro Fuzzy Inference System for CO2 Emissions Data

Abstract

Carbon dioxide (CO2) emission is one of the environmental issues nowadays and reported has been steadily increasing in recent years due to rapid unsustainable economic activities and industrialization. Many intelligence-based models have been employed to investigate CO2 emissions forecast. However, the efficiencies of forecasting models are sometimes very inconclusive. This paper aims to examine the performance of an intelligence-based model, adaptive-neuro fuzzy inference system (ANFIS) model to forecast CO2 emissions data. Twenty years of time series data from the year 1993 to 2013 were used as input and output data to the ANFIS model. The selected input variables are fuel mix, transport, gross domestic product (GDP), and population whereas the output time series data are CO2 emissions. Two statistical measurement errors were computed to examine the performance of the ANFIS. A performance comparison with the possibilistic fuzzy linear regression (PFLR) model was also implemented. The measurement errors indicate that the ANFIS model outperformed the PFLR.