Scholarlink

<p>This study investigates the spatial and temporal variability of Heating Degree Days (HDD) and Cooling Degree Days (CDD) across various regions of Algeria, aiming to assess their impacts on climate-sensitive energy demand and long-term planning. Using descriptive statistics, spatial autocorrelation analysis (Moran’s Index), and trend analysis (Mann-Kendall test, Innovation Trend Analysis), significant warming trends were observed in several Algerian regions, reflected by increasing CDD values and decreasing HDD values, indicating changes in heating and cooling needs. To identify regional patterns, the optimal number of clusters was determined using the Average Silhouette Method, followed by Hierarchical Cluster Analysis (HCA), which revealed two distinct clusters for CDD and four clusters for HDD, highlighting notable geographic disparities in thermal energy requirements. Within each cluster, three supervised machine learning models including Random Forest, AdaBoost, and XGBoost, were trained and evaluated. The XGBoost model consistently outperformed the others across all accuracy metrics, including RMSE, MAE, and R². The results emphasize the importance of integrating spatial classification with advanced machine learning techniques to improve localized energy forecasting, aiming to enhance energy efficiency and strengthen climate change adaptation strategies.</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

Assessing energy stress for energy efficiency under climate variability in Algeria using multivariate statistical analysis and machine learning approaches

  • 期刊论文

  • 发表日期：

    2025年9月27日

查看全文
  • Nassim Bouzid,
  • Karima Aksa,
  • Mohamed Djamel Mouss

摘要

This study investigates the spatial and temporal variability of Heating Degree Days (HDD) and Cooling Degree Days (CDD) across various regions of Algeria, aiming to assess their impacts on climate-sensitive energy demand and long-term planning. Using descriptive statistics, spatial autocorrelation analysis (Moran’s Index), and trend analysis (Mann-Kendall test, Innovation Trend Analysis), significant warming trends were observed in several Algerian regions, reflected by increasing CDD values and decreasing HDD values, indicating changes in heating and cooling needs. To identify regional patterns, the optimal number of clusters was determined using the Average Silhouette Method, followed by Hierarchical Cluster Analysis (HCA), which revealed two distinct clusters for CDD and four clusters for HDD, highlighting notable geographic disparities in thermal energy requirements. Within each cluster, three supervised machine learning models including Random Forest, AdaBoost, and XGBoost, were trained and evaluated. The XGBoost model consistently outperformed the others across all accuracy metrics, including RMSE, MAE, and R². The results emphasize the importance of integrating spatial classification with advanced machine learning techniques to improve localized energy forecasting, aiming to enhance energy efficiency and strengthen climate change adaptation strategies.