Scholarlink

The worldwide change in earth’s temperature, particularly in tropical regions, has drawn the attention of experts due to its impact on human life. The availability of extensive weather observation data and the utilization of advanced data science tools have inspired researchers to investigate trends and minimize errors in massive volumes of data for accurate weather forecasting across various weather parameters. The goal of this analysis is to generate a reliable and flexible statistical model that accurately predicts the temperature of Delhi without any errors. The climate of Delhi is a combination of monsoon-influenced humid subtropical and droughty, characterized by significant fluctuations in both summer and winter temperatures. The prediction model being suggested utilizes Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM) as the hidden layers within the residual network, which highly outperforms in terms of mean square error (MSE) as measure to other cutting-edge algorithms like LSTM, GRU, and ResNet.

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

Statistical Modeling of Temperature Prediction Using Residual Network

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Aakanksha Sharaff,
  • Ankit Sonkusare,
  • Ankita Pal,
  • Sheetal Pahadi

摘要

The worldwide change in earth’s temperature, particularly in tropical regions, has drawn the attention of experts due to its impact on human life. The availability of extensive weather observation data and the utilization of advanced data science tools have inspired researchers to investigate trends and minimize errors in massive volumes of data for accurate weather forecasting across various weather parameters. The goal of this analysis is to generate a reliable and flexible statistical model that accurately predicts the temperature of Delhi without any errors. The climate of Delhi is a combination of monsoon-influenced humid subtropical and droughty, characterized by significant fluctuations in both summer and winter temperatures. The prediction model being suggested utilizes Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM) as the hidden layers within the residual network, which highly outperforms in terms of mean square error (MSE) as measure to other cutting-edge algorithms like LSTM, GRU, and ResNet.