Forecasting CO2 Emissions in Iran Using Important Economic Indicators and Using Deep Learning Models

The increase in greenhouse gas emissions in recent years has caused great concern to many communities and environmentalists; one of these important greenhouse gases is carbon dioxide (CO2). In this study, using important economic variables and indicators and time series data series of 1970-2018, which were divided into 5 separate groups with a set of data, and predicted the amount of CO2 emissions in Iran. For this subject, deep learning models of machine learning subset have been used. It was a multivariate issue and a set of objectives that predicted the amount of CO2 emissions for the next 5 years (5 years after 2018) and finally compared the forecasts for 2019 and 2020 with the actual CO2 of these years to verify the forecasts. The results obtained for all 6 experimental groups show that the amount of CO2 emissions in Iran will follow an upward trend in the coming years and by 2023 the amount of CO2 emissions will reach 850 to 900 million tons, which could be an environmental and dangerous disaster. Be for humans. Therefore, it is suggested that the government use a long-term plan with emphasis on important groups, culture building in the community and the establishment of more specific laws to control the amount of CO2 emissions.