The Prediction of Carbon Dioxide Emissions and Parameter Analysis Based on the LSTM

Yuxi Ji

2023

Abstract

The development of effective climate change mitigation methods and the understanding of the future effects of climate change are made possible by accurate projections of carbon dioxide (CO2) emissions. This paper uses the Long short-term memory (LSTM) model to increase the prediction accuracy of CO2 emissions. Prediction issues can benefit from the use of the LSTM model. Specifically, this paper compares the Mean squared error (MSE) value, representing the precision of CO2 emission prediction, for several LSTM layers and epochs in great detail. This study is conducted on the U.S. Energy Information Administration’s CO2 emissions from the coal power industry dataset. The experiment’s findings show that increasing the number of layers of LSTM can increase the prediction accuracy of CO2 emissions, while reducing the number of layers would decrease that accuracy. Meanwhile, the number of epochs with the maximum prediction accuracy of CO2 emissions under various epochs is 10, and there is no direct relationship between epochs and prediction accuracy. This paper provides an efficient CO2 emission prediction model to provide a practical method to mitigate the greenhouse effect by optimizing the parameters of the LSTM model.

Paper Citation

in Harvard Style

Ji Y. (2023). The Prediction of Carbon Dioxide Emissions and Parameter Analysis Based on the LSTM. In Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML; ISBN 978-989-758-705-4, SciTePress, pages 527-531. DOI: 10.5220/0012800100003885

in Bibtex Style

@conference{daml23,
author={Yuxi Ji},
title={The Prediction of Carbon Dioxide Emissions and Parameter Analysis Based on the LSTM},
booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},
year={2023},
pages={527-531},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012800100003885},
isbn={978-989-758-705-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML
TI - The Prediction of Carbon Dioxide Emissions and Parameter Analysis Based on the LSTM
SN - 978-989-758-705-4
AU - Ji Y.
PY - 2023
SP - 527
EP - 531
DO - 10.5220/0012800100003885
PB - SciTePress