Scholarlink

<p>This study introduces the Hybrid Recurrent Ensemble Depth (HyRED) framework to enhance crop yield prediction in Iraq by integrating diverse environmental and agricultural parameters. The analysis utilises the Valued Agricultural Grounds (VAG) dataset, a novel resource developed exclusively for this study, comprising data from 1034 agricultural plots in Wasit Governorate, Iraq. This unique dataset combines soil properties, agricultural practices, meteorological data, geographic information, and remote sensing indicators such as vegetation health and biomass indices. By using this comprehensive dataset, HyRED employs an advanced hierarchical ensemble structure that integrates memory-based and pattern-recognition modules to capture both short-term temporal patterns and long-term dependencies. This approach achieves high predictive accuracy (MSE = 0.432) while reducing overfitting and enhancing generalisability. By using this novel dataset, remote sensing indicators, and advanced predictive tools, this study offers actionable recommendations for improving agricultural productivity and food security in Iraq, setting a new benchmark for sustainable agriculture research. Notable insights include the identification of optimal fertiliser rates: 60&#xa0;kg / IQ Dunum for diammonium phosphate (DAP) and 100&#xa0;kg / IQ Dunum for nitrogen, which increased productivity by up to 25% compared to conventional practices. Positive correlations with nitrogen (0.8985) and DAP (0.7853) underscore their importance, while negative correlations, such as soil electrical conductivity (−&#xa0;0.9220) and extreme temperatures (−&#xa0;0.5967), highlight environmental challenges. Integrating remote sensing indicators revealed critical growth stages and stress patterns, providing actionable insights to optimise crop management. This study sets a benchmark for sustainable agriculture by combining innovative predictive tools with a novel dataset to improve productivity and food security in Iraq.</p>

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

A hybrid ensemble framework for integrating environmental parameters for improved crop yield prediction and sustainable agricultural decision-making in Iraq

  • 期刊论文

  • 发表日期：

    2025年4月26日

查看全文
  • Ghassan Faisal Albaaji,
  • Vinod Chandra S S

摘要

This study introduces the Hybrid Recurrent Ensemble Depth (HyRED) framework to enhance crop yield prediction in Iraq by integrating diverse environmental and agricultural parameters. The analysis utilises the Valued Agricultural Grounds (VAG) dataset, a novel resource developed exclusively for this study, comprising data from 1034 agricultural plots in Wasit Governorate, Iraq. This unique dataset combines soil properties, agricultural practices, meteorological data, geographic information, and remote sensing indicators such as vegetation health and biomass indices. By using this comprehensive dataset, HyRED employs an advanced hierarchical ensemble structure that integrates memory-based and pattern-recognition modules to capture both short-term temporal patterns and long-term dependencies. This approach achieves high predictive accuracy (MSE = 0.432) while reducing overfitting and enhancing generalisability. By using this novel dataset, remote sensing indicators, and advanced predictive tools, this study offers actionable recommendations for improving agricultural productivity and food security in Iraq, setting a new benchmark for sustainable agriculture research. Notable insights include the identification of optimal fertiliser rates: 60 kg / IQ Dunum for diammonium phosphate (DAP) and 100 kg / IQ Dunum for nitrogen, which increased productivity by up to 25% compared to conventional practices. Positive correlations with nitrogen (0.8985) and DAP (0.7853) underscore their importance, while negative correlations, such as soil electrical conductivity (− 0.9220) and extreme temperatures (− 0.5967), highlight environmental challenges. Integrating remote sensing indicators revealed critical growth stages and stress patterns, providing actionable insights to optimise crop management. This study sets a benchmark for sustainable agriculture by combining innovative predictive tools with a novel dataset to improve productivity and food security in Iraq.