Scholarlink

This study uses machine learningMachine learning models, Random Forest (RF) and XGBoostXGBoost, to predict global temperature anomalies. It examines data from 1983 to 2008 to compare the accuracy of these models. The results show that XGBoostXGBoost performs better, with an R2 of 0.83, RMSE of 0.07 °C, and MAE of 0.06 °C. Random Forest achieves an R2 of 0.73, RMSE of 0.09 °C, and MAE of 0.07 °C. Feature importance analysis shows that the Multivariate El Niño/Southern Oscillation Index (MEI), methane (CH4), and carbon dioxide (CO2) are the most important factors affecting temperature. The study shows that machine learningMachine learning models can capture complex climate patterns and provide accurate predictions. It also provides a framework that can be applied to other regions and climate variables to improve climate research. This work highlights the value of using data-driven approaches to address climate changeClimate change challenges.

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

Harnessing Machine Learning for Climate Prediction: Evaluating the Efficacy of Random Forest and XGBoost Models

  • 章节

  • 发表日期：

    2025年

查看全文
  • Shyam Lochan Bora,
  • Jayanta Das,
  • Shankar Jyoti Nath,
  • Kalyan Bhuyan,
  • Partha Jyoti Hazarika

摘要

This study uses machine learningMachine learning models, Random Forest (RF) and XGBoostXGBoost, to predict global temperature anomalies. It examines data from 1983 to 2008 to compare the accuracy of these models. The results show that XGBoostXGBoost performs better, with an R2 of 0.83, RMSE of 0.07 °C, and MAE of 0.06 °C. Random Forest achieves an R2 of 0.73, RMSE of 0.09 °C, and MAE of 0.07 °C. Feature importance analysis shows that the Multivariate El Niño/Southern Oscillation Index (MEI), methane (CH4), and carbon dioxide (CO2) are the most important factors affecting temperature. The study shows that machine learningMachine learning models can capture complex climate patterns and provide accurate predictions. It also provides a framework that can be applied to other regions and climate variables to improve climate research. This work highlights the value of using data-driven approaches to address climate changeClimate change challenges.