Scholarlink

<p>The exponential growth of global data has accelerated big data research and its applications across various fields. This surge in data, coupled with advancements in information and communication technologies, has enhanced the efficiency of water quality management. This paper explores the influence of big data and machine learning technologies on water resource management and quality monitoring. Dissolved oxygen levels in rivers, a crucial indicator of water quality, require continuous monitoring to support sustainable environmental management. Integrated with the Internet of Things, machine learning techniques enable real-time monitoring and pollution prediction, facilitating the development of low-cost water quality monitoring systems. In this study, we applied the long short-term memory algorithm to predict dissolved oxygen levels in the Oncheon Stream watershed, comparing the accuracy of 1-day and 15-day forecasts. Missing and outliers data collected during the process were addressed using linear interpolation. For the Bugok Bridge site upstream of the Oncheoncheon Stream, one-day predictions using hourly and daily data yielded R² values of 0.8185 and 0.7876, respectively, indicating that the use of hourly data provided superior prediction performance. This research underscores the importance of big data in water quality prediction and demonstrates the potential of machine learning in improving water resource management.</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

Comparative analysis for accuracy of deep learning-based dissolved oxygen prediction model by temporal resolution of learning data: urban stream in Republic of Korea

  • 期刊论文

  • 发表日期：

    2025年9月25日

查看全文
  • Heesung Lim,
  • Hyungjin Shin,
  • Jaenam Lee,
  • Seulgi Lee,
  • Jongwon Do,
  • Isaya Kisekka,
  • Gyumin Lee,
  • Youngkyu Jin

摘要

The exponential growth of global data has accelerated big data research and its applications across various fields. This surge in data, coupled with advancements in information and communication technologies, has enhanced the efficiency of water quality management. This paper explores the influence of big data and machine learning technologies on water resource management and quality monitoring. Dissolved oxygen levels in rivers, a crucial indicator of water quality, require continuous monitoring to support sustainable environmental management. Integrated with the Internet of Things, machine learning techniques enable real-time monitoring and pollution prediction, facilitating the development of low-cost water quality monitoring systems. In this study, we applied the long short-term memory algorithm to predict dissolved oxygen levels in the Oncheon Stream watershed, comparing the accuracy of 1-day and 15-day forecasts. Missing and outliers data collected during the process were addressed using linear interpolation. For the Bugok Bridge site upstream of the Oncheoncheon Stream, one-day predictions using hourly and daily data yielded R² values of 0.8185 and 0.7876, respectively, indicating that the use of hourly data provided superior prediction performance. This research underscores the importance of big data in water quality prediction and demonstrates the potential of machine learning in improving water resource management.