Open AccessTechnological assessment and modeling of energy‐related CO 2 emissions for the G8 countries by using hybrid IWO algorithm based on SVM
Author(s) -
Ghazvini Mahyar,
Dehghani Madvar Mohammad,
Ahmadi Mohammad Hossein,
Rezaei Mohammad Hossein,
El Haj Assad Mamdouh,
Nabipour Narjes,
Kumar Ravinder
Publication year - 2020
Publication title -
energy science and engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.638
H-Index - 29
ISSN - 2050-0505
DOI - 10.1002/ese3.593
Subject(s) - support vector machine , artificial neural network , test data , section (typography) , renewable energy , algorithm , engineering , energy (signal processing) , computer science , simulation , artificial intelligence , machine learning , statistics , mathematics , software engineering , electrical engineering , operating system
Abstract Recently, energy‐related CO 2 emissions are considered as one of the most crucial issues and are promptly augmented due to further urbanization. In this paper, in order to model and calculate yearly CO 2 emission, an artificial neural network is used. For the first time, the IWO‐SVM method has been applied in modeling energy‐related CO 2 emissions. In this regard, consumption of different energy sources such as renewable energy, natural gas, coal, and oil, and GDP of the G8 countries in various years (from 1990 to 2016) are regarded as input in the present study. For the aim of evaluating the exact ability of the SVM and SVM‐IWO models, the performance of these models in three different modes is carried out on the basis of the number of data in the test and train sections. For this purpose, implementations are split into three categories ( a = 80% of the data for the train section and 20% for the test section; b = 70% of the data for the train section and 30% for the test section; and c = 60% of the data for the train section and 40% for the test section). Furthermore, five scenarios were selected on the basis of the number of input parameters and input parameters for achieving the best model. As indicated in the results in all scenarios, the correlation of the model with the hybrid invasive weed algorithm based on SVM is more favorable than that in the support vector machine model, due to better training of the SVM‐IWO model than the SVM model. Moreover, the technological orientations of the G8 countries to mitigate CO 2 emissions are determined through patent analysis. While the patents have essential information, investigating the published patent by a country could be helpful for determination of technological orientations. Hence, all published patents by these countries are extracted and deeply analyzed. In the next step, to find out main technological approaches, all patents and their intents have been studied. Eventually, the technological life cycles and trends of each main technology are drawn.