Artificial intelligence applications in greenhouse gas emission calculations

Greenhouse gases that trigger global warming stem from both natural factors and human activities. The emission of greenhouse gases resulting from the use of fossil fuels constitutes a paramount factor in global warming. Particularly, carbon dioxide exerts the most potent impact on global warming, as it acts as a heat-absorbing gas, with its effect being considerably significant. Global agreements, notably the Paris Agreement, have taken significant steps towards reducing human activities and embracing net zero emission targets. Consequently, it is anticipated that all countries will strive to achieve their greenhouse gas emission reduction goals by implementing sustainable and pragmatic programs. Machine learning methods have been employed in this study, utilizing financial, economic, and human development indicators, population data, deforestation rate and energy consumption data to calculate future greenhouse gas emission levels in certain countries. In this study, comparisons have been made using deep learning techniques, such as Long Short-Term Memory (LSTM) and hybrid CNN-RNN models, through the MATLAB program to achieve the objective of reducing greenhouse gas emissions. Additionally, future greenhouse gas emission predictions were made by comparing LSTM modeling results with those obtained through NARX modeling for time-series data. This study is also expected to facilitate the development of sustainable programs by considering different data for countries to achieve their greenhouse gas emission reduction targets.