Scholarlink

Meteorological parameters (air temperature, dew point temperature, wind speed, relative humidity, atmospheric pressure, precipitation intensity) and the pollution factor (the amount of coal consumed by the Bishkek city’s thermal power plant) are considered as predictor variables in problems of forecasting atmospheric air pollution based on machine learning methods. Cross correlations of PM2.5 particulate matter concentrations and predictor variables were studied, taking into account the lag effect for winter and summer seasons from 09.02.2019 to 14.09.2023. The most significant Pearson correlation coefficients and their corresponding lags are shown. The results obtained can be useful for understanding air pollution processes and construct more accurate forecasting models than those currently available, based on machine learning methods, with the potential impact on air quality management.

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

Analysis of Predictive Variables for Air Pollution Forecasting Models Based on Machine Learning Methods

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • N. M. Lychenko

摘要

Meteorological parameters (air temperature, dew point temperature, wind speed, relative humidity, atmospheric pressure, precipitation intensity) and the pollution factor (the amount of coal consumed by the Bishkek city’s thermal power plant) are considered as predictor variables in problems of forecasting atmospheric air pollution based on machine learning methods. Cross correlations of PM2.5 particulate matter concentrations and predictor variables were studied, taking into account the lag effect for winter and summer seasons from 09.02.2019 to 14.09.2023. The most significant Pearson correlation coefficients and their corresponding lags are shown. The results obtained can be useful for understanding air pollution processes and construct more accurate forecasting models than those currently available, based on machine learning methods, with the potential impact on air quality management.