Published Online:January 2025
Product Name:The IUP Journal of Applied Economics
Product Type:Article
Product Code:IJAE020125
DOI:10.71329/IUPJAE/2025.16.1.24-47
Author Name:Vengalarao Pachava and Aparna A
Availability:YES
Subject/Domain:Economics
Download Format:PDF
Pages:24-47
This study explores CO2 emissions forecasting in India and South Africa, two major contributors among BRICS nations, using machine learning (ML) and deep learning (DL) techniques such as Support Vector Regression (SVR), Random Forest (RF), and Long Short-Term Memory (LSTM). Analyzing data from 1992 to 2021 on GDP, population, industry value added, and domestic credit, the study evaluates model performance using metrics such as Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results reveal that LSTM outperforms SVR and RF in accuracy and in capturing dynamic temporal trends, with forecasts indicating a steady rise in emissions through 2030. Key drivers differ between countries, with population and GDP dominating in India, and GDP and population leading in South Africa. The study highlights LSTM’s potential for dynamic emissions forecasting, offering data-driven insights for policymakers to design sustainable growth strategies, while emphasizing the need for future research integrating additional variables, hybrid models, and real-time data for improved forecasting.
BRICS nations (Brazil, Russia, India, China, and South Africa) have been major drivers of global economic growth over the last two decades. These countries account for over 40% of the global population and contribute approximately 32% of the world’s GDP. However, this rapid economic expansion has been coupled with high energy consumption, leading to a substantial environmental footprint. BRICS nations alone contributed 43.3% to global carbon emissions in 2020, highlighting the environmental cost of their growth trajectory (Statista, 2023).