We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Previsão do preço do carbono por modelos de aprendizado de máquina.
- Authors
Yoshio Kanda, Jorge; de Leon Ferreira de Carvalho, André Carlos Ponce
- Abstract
Controlling and reducing greenhouse gas emissions are necessary actions to avoid the possible consequences of climate change. In this context, the carbon market is of great relevant, especially for developing countries, such like Brazil, which has an immense environmental wealth with its Amazon Forest. The purpose of this study is to identify sustainability datasets to be used by predictive machine learning (ML) models that are able to accurately estimate the price of carbon practiced in the main world market. In our computational experiments, algorithms were implemented from different ML algorithms, using different datasets as input parameters. The results obtained show that Amazonian data seem to have a direct relationship with the price of carbon practiced in the world market. A feature selection procedure was applied to the union of the Amazonian datasets and submitted to the same ML models to verify if there are improvements in the predictive performance. Therefore, with an accurate estimate of the carbon price and Brazil regulating the rules for carbon trading, the Amazon Region tends to benefit from significant gains in environmental, economic, and social aspects.
- Subjects
CARBON credits; GREENHOUSE gas mitigation; ARTIFICIAL intelligence; MACHINE learning; CARBON pricing; REGRESSION analysis; FEATURE selection; SUSTAINABILITY; CARBON offsetting
- Publication
Amazônia, Organizações e Sustentabilidade (AOS), 2023, Vol 12, Issue 2, p158
- ISSN
2238-8893
- Publication type
Article
- DOI
10.17648/aos.v12i2.2916