Efficiency of simulation and machine learning algorithms for modeling and forecasting greenhouse gas emissions: A review
Authors: Olorunsola, O., Adedoyin, F. and Adebiyi, A.
Pages: 161-181
DOI: 10.1016/B978-0-443-33971-4.00005-2
Abstract:The use of Simulation and Machine Learning Algorithms for modeling and Forecasting Greenhouse Gas (GHG) Emissions has been pronounced in the literature; however, there is no documentation or classification of the various categories across time which is the focus of this paper. This paper reviews research in this area while highlighting Artificial neural networks (ANNs) and support vector machine (SVM) which are popular machine learning algorithms used in forecasting. The core finding of this paper is that the combination of both algorithms can increase accuracy and robustness of forecasting. Also, in this research, ANN and SVM algorithms are found to be combined to produce high accuracy. The ideology behind this combination is to utilize the strengths of both algorithms as ANN is great at complex nonlinear relationships between input and output; on the other hand, SVM is useful in finding a linear or nonlinear boundary between two classes. However, ANNs are prone to overfitting and require large amounts of data training and SVM may underperform when dealing with complex data.
Source: Scopus