Scholarlink

To achieve the sustainable development of the regional economy in Xiangxi Prefecture, it is necessary to predict the production of the tea industry in Xiangxi Prefecture to assist the government in decision-making, as well as help farmers and enterprises adjust their planting and marketing strategies. Meteorological, production, and economic factors influence tea production in western Hunan, making in-depth prediction and study challenging. In this study, we use the random forest and gradient enhancement model to construct a multi-dimensional influencing factor model to predict the tea production. The test results indicate that the combined model predicts better than the individual model, with the R2 score going up from 0.892 to 0.973 in the single random forest model and the root mean square error decreasing by about 52.56%. The prediction model proposed in this paper has better prediction accuracy than the traditional tea model and has important reference value in predicting tea yield.

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

Investigation of Tea Yield Forecasting in Xiangxi Prefecture Utilizing Random Forest and Gradient Boosting Techniques

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Bingjie Wang,
  • Xinyi Tao,
  • Jingwen Li,
  • Pengfei Yin,
  • Junping Shi,
  • Fanghui Mo,
  • Qian Xu

摘要

To achieve the sustainable development of the regional economy in Xiangxi Prefecture, it is necessary to predict the production of the tea industry in Xiangxi Prefecture to assist the government in decision-making, as well as help farmers and enterprises adjust their planting and marketing strategies. Meteorological, production, and economic factors influence tea production in western Hunan, making in-depth prediction and study challenging. In this study, we use the random forest and gradient enhancement model to construct a multi-dimensional influencing factor model to predict the tea production. The test results indicate that the combined model predicts better than the individual model, with the R2 score going up from 0.892 to 0.973 in the single random forest model and the root mean square error decreasing by about 52.56%. The prediction model proposed in this paper has better prediction accuracy than the traditional tea model and has important reference value in predicting tea yield.