Scholarlink

Rainfall prediction (RP) utilizing machine learning (ML) is an imperative tool for agricultural and economic practices. RP is also considered as a challenging chore in weather forecasting. An accurate RP can assist farmers to make knowledgeable decisions regarding crop planting, harvesting and fertilization. It also helps in the prevention of floods, disaster management and water resource management. Therefore, the prediction of precise rainfall is significant for diverse applications of a country. Here, Adaptive Lotus Effect algorithm-based Long Short-Term Memory (ALEA_LSTM) is presented for RP utilizing time series data (TSD). Initially, TSD is obtained from a specific database and thereafter technical indicators for RP are extracted from the considered input TSD. Afterwards, feature selection (FS) is accomplished to choose suitable features employing mutual information (MI). Then, data augmentation (DA) is conducted by the oversampling method to augment the dimensionality of data. At last, RP is executed by Long Short-Term Memory (LSTM) that is trained by the Adaptive Lotus Effect Algorithm (ALEA). Moreover, ALEA is devised by integrating the adaptive concept with the Lotus Effect Algorithm (LEA). In addition, ALEA_LSTM achieved minimal values of Mean absolute percentage error (MAPE) of about 0.359, Mean squared error (MSE) of about 0.133 and Root mean squared error (RMSE) of about 0.364.

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

CFALEA_LSTM: Adaptive Lotus Effect Algorithm Enabled Long Short-Term Memory for Rainfall Prediction Using Time Series Data

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • P. Vijaya,
  • Satish Chander,
  • Roshan Fernandes,
  • Anisha P. Rodrigues,
  • B. Supriya

摘要

Rainfall prediction (RP) utilizing machine learning (ML) is an imperative tool for agricultural and economic practices. RP is also considered as a challenging chore in weather forecasting. An accurate RP can assist farmers to make knowledgeable decisions regarding crop planting, harvesting and fertilization. It also helps in the prevention of floods, disaster management and water resource management. Therefore, the prediction of precise rainfall is significant for diverse applications of a country. Here, Adaptive Lotus Effect algorithm-based Long Short-Term Memory (ALEA_LSTM) is presented for RP utilizing time series data (TSD). Initially, TSD is obtained from a specific database and thereafter technical indicators for RP are extracted from the considered input TSD. Afterwards, feature selection (FS) is accomplished to choose suitable features employing mutual information (MI). Then, data augmentation (DA) is conducted by the oversampling method to augment the dimensionality of data. At last, RP is executed by Long Short-Term Memory (LSTM) that is trained by the Adaptive Lotus Effect Algorithm (ALEA). Moreover, ALEA is devised by integrating the adaptive concept with the Lotus Effect Algorithm (LEA). In addition, ALEA_LSTM achieved minimal values of Mean absolute percentage error (MAPE) of about 0.359, Mean squared error (MSE) of about 0.133 and Root mean squared error (RMSE) of about 0.364.