| Titre : | Institutional quality and CO² emissions : Evidence from explainable artificial intelligence modeling |
| Auteurs : | Nicolae STEF, Auteur ; Samir BEN JABEUR, Auteur ; Hakan BASAGAOGLU, Auteur ; Debaditya CHAKRABORTY, Auteur |
| Type de document : | Article |
| Année de publication : | 2022 |
| Importance matérielle : | 26 p. |
| Langues: | Anglais |
| Concepts : | Finance, Governance, SER |
| Mots-clés : |
[Aciège 2017] GESTION DE LA QUALITE [Aciège 2017] INTELLIGENCE ARTIFICIELLE [Aciège 2017] THEORIE DES JEUX [Aciège 2017] METHODE |
| Résumé : | This study examines how the institutional quality ranks in the order of importance of variables in explaining global carbon emissions. Our predictive modeling approach relies on explainable artificial intelligence (XAI) models involving random forest, extremely randomized trees, extreme gradient boosting, and light gradient boosting that are combined with the game theory-based Shapley method. Using data from 136 countries for the period from 1996 to 2016, we identified the extremely randomized tree as the best-performing XAI model with the second highest prediction accuracy and the best explainability with causality. According to that model, the most important institutional dimensions were identified to be the rule of law (ranked 4th) followed by citizens’ political voice (ranked 7th), corruption control (ranked 8th), and government effectiveness (ranked 10th). The XAI modeling results further revealed that when such institutional features are effective, the best strategy to reduce carbon emissions would be to raise the annual GDP per capita (ranked 1st) above $ 10,000 and the consumption of renewable energy (ranked 2nd) above 24% of the total energy consumption. |
| Citation norme APA : | Stef, N., Basagaoglu, H., Chakraborty, D., & ben Jabeur, S. (2022). Institutional quality and CO2 emissions : Evidence from explainable artificial intelligence modeling. 4th Edition of Ethical Finance and Sustainability Conference (EFS-2022), 1‑26. |
| Document numérique : | Pour voir le document numérique, pensez à vous identifier |
Documents numériques (1)
 AI and Innovation_13_04_2022. Nicolae STEF Adobe Acrobat PDF |
