期刊名称：International Journal of Advances in Engineering and Management
电子版ISSN：2395-5252
出版年度：2021
卷号：3
期号：8
页码：799-808
DOI：10.35629/5252-0308638644
语种：English
出版社：IJAEM JOURNAL
摘要：African countries are the main of the importers of used vehicles from Japan and Europe, some countries install the regulations for limitation of poorest quality vehicles, and such regulations are often poorly enforced, due to corruption. The old cars have higher CO2 emissions and are dumped on us because they are no longer considered as fit for the roads in their countries of origin, many of those used cars are being used for commercial, public transportation, and are causing roads accident which increases deaths rate in African roads as different consulted reports stated. To understand the relationship between the CO2 emissions, vehicles imported annually, the amount of fuel consumed, and other parameters that make the huge volume of data need advanced methods such as machine learning techniques in data analysis to showcase the issue in Eastern, Southern and Western African countries and then predict the emission of CO2 in these countries in case the situation remains the same as it is nowadays which is the main objective of this study. In this study, we are mainly aim at assessing the past and current, and predicting CO2 emissions in Africa. The output indicated that the LR has the ability of predicting the CO2 emission from solid fuel with MAPE with meaningful error rates at 1.024% and 1.901% for predicting AES and USA respectively and when predicting CO2 emission from liquid fuel with LR, findings present the 2.36% and 2.85% for MAE when predicting Japan and USA emissions; For solid fuel are well predicted by SVM with MAPE at 1.28%, 4.28%, and 1.43% consecutively for AES, AWC, and EU.
