Search

To improve the school's teaching plan, optimize the online learning system, and help students achieve better learning outcomes, an educative data mining model for the supervision of the e-learning process was established. Statistical analysis and visualization in data mining techniques, association rule algorithms, and clustering algorithms were applied. The teaching data of a college English teaching management platform was systematically analyzed. A related conclusion was drawn on the relationship between students' English learning effects and online learning habits. The results showed that this method could effectively help teachers judge students' online learning results, understand their online learning status, and improve their online learning process. Therefore, the model can improve the effectiveness of students' online learning. [ABSTRACT FROM AUTHOR]

• The proposed multi-factor-based forecasting model outperforms ARIMA, ARMAX and SVM. • Energy, economy and industry are main influencing factors on carbon emissions. • With no constraint, carbon emissions of China in 2020 will be 6078.21 million tonnes. • Carbon intensity in the middle region will be the most serious in 2023. Accurate carbon emissions forecasting plays a pivotal role in reducing global warming by providing references to formulate emission reduction policies. Although numerous studies have focused on forecasting China's carbon emissions, the results of different methods are contradicting, because they are based on different data and use different parameters. This paper aims to propose a hybrid carbon emissions forecasting model based on multi-factor identification to offer reliable forecasting results. First, association rule algorithm was applied to find influencing factors and analyse their joint effects on carbon emissions from the perspective of time and space. Energy consumption, economic growth, industrial structure, foreign direct investment, and urbanization are proven to be the five major factors that can cause an increase in carbon emissions. Second, a multivariate grey model optimized by firefly algorithm was utilized to conduct carbon emissions forecasting under different scenarios. Empirical results indicated that the proposed hybrid model had the best performance compared to other methods. If no effective measures are taken, it is difficult for China to realize its goal for carbon emissions reduction in 2020. [ABSTRACT FROM AUTHOR]