Scholarlink

The development of long-term rainfall forecasting presented n this research used comparative deep-learning approaches. The long-term meteorological data is collected from the National Meteorological Agency of Ethiopia at the Jimma service center in the southwestern part of the Oromia region. Using data collected during the last 15 years, missing values are pre-processed using the anticipated algorithm, and the data are further normalized to enhance the effectiveness of the forecasting system. The results of CNN and LSTM models are compared with different performance measurements, such as root mean absolute error and mean absolute error. The CNN algorithm’s MAE and RMSE values were 3.58 and 7.85, while the LSTM algorithm’s MAE and RMSE values were 4.97 and 8.61. The CNN algorithm was able to forecast an accuracy of 98.18%. On the other hand, LSTM algorithms achieved an accuracy of 96.73%. As a result, CNN performed well in terms of accuracy when compared to the LSTM method.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Long Term Rainfall Forecasting Using Deep Learning Approach: The Case of Ilu Abba Borzone

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Alemayehu Etana Duguma,
  • Ramata Mosissa Gichila,
  • Wagari Goje Tuji

摘要

The development of long-term rainfall forecasting presented n this research used comparative deep-learning approaches. The long-term meteorological data is collected from the National Meteorological Agency of Ethiopia at the Jimma service center in the southwestern part of the Oromia region. Using data collected during the last 15 years, missing values are pre-processed using the anticipated algorithm, and the data are further normalized to enhance the effectiveness of the forecasting system. The results of CNN and LSTM models are compared with different performance measurements, such as root mean absolute error and mean absolute error. The CNN algorithm’s MAE and RMSE values were 3.58 and 7.85, while the LSTM algorithm’s MAE and RMSE values were 4.97 and 8.61. The CNN algorithm was able to forecast an accuracy of 98.18%. On the other hand, LSTM algorithms achieved an accuracy of 96.73%. As a result, CNN performed well in terms of accuracy when compared to the LSTM method.