Application of LSTM Machine Learning to Prediction Precipitation in Beijing Area

Jincheng Yuan, Zhongmu Li

2022

Abstract

Machine learning plays a vital role in climate prediction. In this study, we apply a mixed model (EEMD-LSTM) combining Ensemble Empirical Mode Decomposition (EEMD) and Long Short-Term Memory Network (LSTM) to predict precipitation in Beijing. EEMD divides the input precipitation data into multiple subseries. We use the LSTM to predict the subsequences separately and combine the forecasting of each subsequence to acquire the end results. By establishing the conventional LSTM model, EEMD-BP model, and BP model for comparison, Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination () are used as evaluation indicators. The RMSE, MAE, and of the EEMD-LSTM model are 1.3337mm, 0.7221mm, and 0.8170, respectively. The EEMD-LSTM model optimizes RMSE by 53.22% and MAE by 56.72% comparing to the conventional LSTM model. It optimizes RMSE by 15.28% and MAE by 16.20% comparing to the EEMD-BP model and optimizes RMSE by 54.10% and MAE by 53.72% comparing to the BP model. of EEMD-LSTM is closest to 1. The study outcomes confirm that the EEMD-LSTM model can obtain the precipitation forecast result of Beijing with higher prediction accuracy.

Paper Citation

in Harvard Style

Yuan J. and Li Z. (2022). Application of LSTM Machine Learning to Prediction Precipitation in Beijing Area. In Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC; ISBN 978-989-758-622-4, SciTePress, pages 191-195. DOI: 10.5220/0011917500003612

in Bibtex Style

@conference{isaic22,
author={Jincheng Yuan and Zhongmu Li},
title={Application of LSTM Machine Learning to Prediction Precipitation in Beijing Area},
booktitle={Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC},
year={2022},
pages={191-195},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011917500003612},
isbn={978-989-758-622-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 3rd International Symposium on Automation, Information and Computing - Volume 1: ISAIC
TI - Application of LSTM Machine Learning to Prediction Precipitation in Beijing Area
SN - 978-989-758-622-4
AU - Yuan J.
AU - Li Z.
PY - 2022
SP - 191
EP - 195
DO - 10.5220/0011917500003612
PB - SciTePress