Scholarlink

Background <p>Urban fires pose significant risks, resulting in substantial financial and human losses. Effective management of firefighting operations is crucial, where response times play a vital role in minimizing casualties and damages. This study aimed to utilize machine learning (ML) algorithms to accurately predict the operation time of firefighting in urban areas and to identify the most critical factors influencing this duration.</p> Materials and method <p>Data from 1402 firefighting operations in Ilam City, collected in 2022, were analyzed. After data preprocessing, which involved checking for duplicates and outlier detection, 1399 incidents were included for analysis. A total of 31 variables were initially examined, with 19 selected for the final models.</p> Results <p>The Generalized Linear Models (GLM) model demonstrated the lowest Root Mean Squared Error (RMSE) (95% CI: 14.21,  ±1.65), indicating superior performance in estimating firefighting operation time. Key predictors identified included the number of personnel, rescue equipment, automobile fire engines, and manual extinguishers. GLM estimated the average operation time at 26.964&#xa0;min (95% CI: 12.75–41.171).</p> Conclusion <p>The study highlights the effectiveness of ML in predicting firefighting operation times and emphasizes the importance of critical resources like personnel and equipment in achieving quicker response times. The findings have significant implications for urban fire management strategies aimed at reducing human and financial losses. Further research should enhance data collection efforts to improve prediction accuracy in firefighting operations.</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

Predicting firefighting operation time in urban areas using machine learning: identifying key determinants for improved emergency response

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月21日

查看全文
  • Ali Sahebi,
  • Hojjat Sayyadi,
  • Bahram Havasy,
  • Yousef Veisani

摘要

Background

Urban fires pose significant risks, resulting in substantial financial and human losses. Effective management of firefighting operations is crucial, where response times play a vital role in minimizing casualties and damages. This study aimed to utilize machine learning (ML) algorithms to accurately predict the operation time of firefighting in urban areas and to identify the most critical factors influencing this duration.

Materials and method

Data from 1402 firefighting operations in Ilam City, collected in 2022, were analyzed. After data preprocessing, which involved checking for duplicates and outlier detection, 1399 incidents were included for analysis. A total of 31 variables were initially examined, with 19 selected for the final models.

Results

The Generalized Linear Models (GLM) model demonstrated the lowest Root Mean Squared Error (RMSE) (95% CI: 14.21,  ±1.65), indicating superior performance in estimating firefighting operation time. Key predictors identified included the number of personnel, rescue equipment, automobile fire engines, and manual extinguishers. GLM estimated the average operation time at 26.964 min (95% CI: 12.75–41.171).

Conclusion

The study highlights the effectiveness of ML in predicting firefighting operation times and emphasizes the importance of critical resources like personnel and equipment in achieving quicker response times. The findings have significant implications for urban fire management strategies aimed at reducing human and financial losses. Further research should enhance data collection efforts to improve prediction accuracy in firefighting operations.